Wan Design Guide

ProCurve Networking by HP
WAN Design Guide

The Lower Layers
August 2005
Introduction ............................................................................................................... 5
Secure WAN Design Overview .................................................................................... 5
Understanding the Customer Requirement ................................................................ 5
Overview of WAN Environments .............................................................................. 7
What is a WAN? .................................................................................................... 7
How is LAN different from WAN?.............................................................................. 9
Types of WAN Circuits............................................................................................. 10
Dedicated Physical Circuits ................................................................................... 10
Switched Physical Circuits .................................................................................... 10
Permanent Virtual Circuits (PVCs).......................................................................... 11
Switched Virtual Circuits (SVCs) ............................................................................ 11
How is LAN similar to WAN?.................................................................................. 12
Designing the Physical and Data Link Layers ................................................................. 15
An Overview of the Local Loop (The Transmission Technologies) ...................................... 15
Executive Summary ............................................................................................ 15
Overview ........................................................................................................... 15
T1 and E1 Technologies .......................................................................................... 16
Summary of Major Points ..................................................................................... 17
How This Technology is Used ................................................................................ 17
Advantages ........................................................................................................ 17
Disadvantages .................................................................................................... 17
What to Determine During Planning or for Implementation........................................ 17
ADSL Technology: .................................................................................................. 18
Summary of Major Points for ADSL ........................................................................ 22
Advantages ........................................................................................................ 23
Disadvantages .................................................................................................... 23
What to Determine During Planning ....................................................................... 23
ISDN .................................................................................................................... 23
ISDN Equipment at the Subscribers Premises ............................................................... 25
Network Termination 1 ........................................................................................ 25
Terminal Equipment ............................................................................................ 25
U Interface......................................................................................................... 26
T Interface ......................................................................................................... 26
S Interface......................................................................................................... 26
R Interface......................................................................................................... 26
Connectors......................................................................................................... 26
Advantages ........................................................................................................ 27
Disadvantages .................................................................................................... 27
Data Link Layer Protocols in the WAN (The Transport Technologies) ................................. 27
Overview ........................................................................................................... 28
HDLC ................................................................................................................... 28
Advantages ........................................................................................................ 29
Disadvantages .................................................................................................... 29
PPP ...................................................................................................................... 29
Advantages ........................................................................................................ 29
Disadvantages .................................................................................................... 30
Frame Relay .......................................................................................................... 30
Advantages ........................................................................................................ 31
Disadvantages .................................................................................................... 31
ATM Technology..................................................................................................... 32
Service Categories .............................................................................................. 32
Definitions ......................................................................................................... 33
Services ............................................................................................................ 33
Traffic Parameters............................................................................................... 33
Service Categories .............................................................................................. 34
Advantages ........................................................................................................ 35
Disadvantages .................................................................................................... 36
Using ATM and DSL together................................................................................. 36

PPPoA (Point to Point Protocol over ATM) Planning Questions..................................... 36
PPPoE (Point to Point Protocol over Ethernet) Planning Questions ............................... 36
RFC 1483 Planning Questions................................................................................ 36
Further Information............................................................................................. 37
Solution Examples for Layer 1 and 2 ......................................................................... 37
PPP Solution Example .......................................................................................... 37
MLPPP Solution Example ...................................................................................... 40
Frame Relay Solution Examples............................................................................. 42
Frame Relay Solution Example 1 ........................................................................... 42
Cisco 2621xm..................................................................................................... 47
Multilink Frame Relay Solution Example.................................................................. 51
ADSL Solution Example ........................................................................................... 52
ADSL Solution Example........................................................................................ 52
ADSL Solution Example with PPPoE over ATM over ADSL........................................... 52
Feature Set Consideration for ProCurve Secure Router 7000dl Series .......................... 55
How to Use the ProCurve Secure Router 7000dl Series ............................................. 55
An Overview of IP, Static and Dynamic Routing Protocols ............................................. 56
Overview ........................................................................................................... 56
IP General............................................................................................................. 56
IP networks and subnetworks ............................................................................... 57
Classful vs. classless IP addressing ........................................................................ 57
IP routing .......................................................................................................... 58
Remote networks ................................................................................................ 58
Routing Information Protocol (RIP) ........................................................................ 58
Static routes....................................................................................................... 58
Administrative distance ........................................................................................ 58
Default static route.............................................................................................. 58
Open Shortest Path First (OSPF)............................................................................ 59
OSPF hierarchy ................................................................................................... 59
The flow of link state information .......................................................................... 60
Multiple OSPF areas............................................................................................. 60
Autonomous System Boundary Router (ASBR)......................................................... 60
Normal areas and stub areas ................................................................................ 61
Not-so-stubby areas (NSSA) ................................................................................. 61
Border Gateway Protocol ......................................................................................... 61
External BGP operation ........................................................................................ 62
Overlapping address spaces and the longest mask rule............................................. 68
How These Technologies are Used ......................................................................... 69
Advantages ........................................................................................................ 69
Disadvantages .................................................................................................... 70
What to Determine During Planning or for Implementation........................................ 70
Solution Examples for Layer 3.................................................................................. 70
OSPF Solution Example 1 ..................................................................................... 70
OSPF Solution Example 2 ..................................................................................... 73
BGP Solution Example 1 ....................................................................................... 80
Feature Set Consideration for ProCurve Secure Router 7000dl Series .......................... 84
How to Use the ProCurve Secure Router 7000dl Series ............................................. 84
Additional Topics ....................................................................................................... 84
High Availability and Redundancy ............................................................................. 84
IP Multicast ........................................................................................................... 85
Security, Access Control Lists, and Virtual Private Networks ......................................... 85
Quality of Service................................................................................................... 85
Appendix A Route summarization.............................................................................. 86
Routing among locations ......................................................................................... 86
Dynamic route exchange ......................................................................................... 86
Network summarization .......................................................................................... 87
Summarization of address space using static routes .................................................... 87
Summarizing remote address space .......................................................................... 88
Summarizing at classful network boundaries .............................................................. 89
Summarizing a larger address space ......................................................................... 89
Appendix B Glossary ............................................................................................... 90
Appendix C Differences with Cisco Routers ................................................................105
Appendix D References...........................................................................................106
Introduction
Since the dawn of time we have had the need to communicate at a distance. From the fleet-footed
messenger running between small villages to the dawn of the telegraph and telephone the goals have
been the same. Bring the message from point A to point B as quickly as possible, with accuracy in
transmission. Later requirements often included cost, but it is not much of a stretch to believe that
villages didnt like loosing fast runners to injury either!
The messages between point A and point B may have changed and become more sophisticated but
the ideas are the same. Today companies that have multiple offices need a cost-effective, efficient
means to exchange data between those offices. Many companies have created intranets or extranets,
which enable customers at different locations to view information and to upload and download
information.
Security is also an issue because often times today the customers intranet (internal network) must be
connected to the Internet in order to conduct business outside their own company and allow the
resources of the Internet in. The various customer location connected through the Internet must be
protected by firewalls.
With these considerations to serve us, this paper will stay focused on what are the basic requirements are
for the transmission of data and site to site communication. At times we will discuss cost considerations
but only at a high level. Reference text books are mentioned in the back of this guide for those requiring
further information. We will introduce various example networks to consider using the ProCurve Secure
Router 7000dl series.
The intended audience for this guide is the technical consultant, with moderate experience in wide area
networking, familiar with the lower four layers of the OSI model, yet who may not be familiar with the
ProCurve Secure Router 7000dl series or an expert in WAN technologies. In short the consultant who has
been focused on LAN networking but has limited experience with WAN. Even though this is the audience
focus, this guide will still be useful at many points even for the most experienced, as it does contain some
configuration comparisons between some of the ProCurve and Cisco routers and other relevant
information for those more experienced.
This guide will also focus most of its attention on the traditional WAN layers, the physical and data link
layers. There are sections discussing the network layer and Internet connectivity, yet considerations at
this point will be limited to technology overviews, and configuration examples, with a focus on the
ProCurve Secure Router 7000dl series. Design considerations for IP addressing, or consideration of the
implications of one dynamic routing protocol over another will not be a major focus of this paper. Never
the less there will be references to direct the reader to further information.
Given this scope the main body of this guide will investigate two major domains of wide area network
design; Designing the Physical and Data Link Layers, Designing the Network Layer and Internet.
Other highlights within each section are:
Overview of Technologies
Advantages and Disadvantages
Summary of Major Points
What to Determine During Planning or
How The Technology is Used
for Implementation
Solution Configuration Examples
Secure WAN Design Overview

Understanding the Customer Requirement
How does one progress from conception to reality in designing their network? This is as difficult to find a
single best answer for as if one were to try determining the best earth-bound route to go from Bannock
to Frankfurt. No single right answer would be given. But knowing this should not keep you from
gathering enough information to make intelligent choices and concessions.
In all cases the reader must remember to work toward a wide area network that is as fast as possible,
within any understood constraints, that handles the data accurately and securely, for a reasonable cost.
To accomplish this we need to ask some basic questions. Below are some suggested questions to help
you get started:
Is this a new installation or replacing an existing?

If existing, what problems does the customer currently face that they would like corrected?
What are the requirements?
What is the rate of data transfer?
Must the network be high speed in both directions or only one?
By answering some of these we might determine performance and reliability requirements. Other
questions would be:
What levels of security need to be in place?

If this customer wants these routers hooked to the Internet then they need to have designed
for the stateful firewall and access policies to be in place.
What would be the areas we could compromise?

Is high performance is priority at all locations or only at some.
Are there areas that are not top priority?
Other questions would be:
Is this for data only or is this a combined data and voice solution?
Is the voice solution traditional TDM voice or Voice over IP?
Does the customer need a strategy for redundancy?
If yes then consider, through risk assessment, what should be redundant the WAN links, the
carrier or ISP, the routers themselves, the modules in the routers?
How much monthly cost do I tolerate for one day of downtime?

Then what would be the best possible network design?
In order to answer this question the previous questions need to be addressed and then
combined with what is available in the locations of the customer locations. Your best possible
design will be a combination of requirements aligned with the technologies and services
available to accomplish them.
These are just a few of the questions one would need to answer before designing the network.
Finally, as mentioned previously, there are many good reference texts available for your use. If the
reader requires further guidance there is a reference section at the end of this paper along with a brief
description of the publication to help you determine if you want to purchase the reference or not. There
are times when smaller networks and upgrades can be designed with estimates, at other times a much
more systematic approach is required. Some would argue as Robert Cahn who writes, It is not possible
to design networks at any scale without algorithms. Yet he also admits, Design problems are much
too hard to be solved exactly. So although this paper does agree that their needs to be objective
analysis and design utilizing tools that are available, its philosophy holds that solid understanding of
network communications, along with knowing the basic infrastructure and cost, a designer can come up
with a reasonably good design without dependency upon algorithms in every design. To those who have
been competent in LAN design can learn WAN technologies and adapt many of them to their current
understanding.
This paper will equip the technical consultant with the major tools required for designing wide area
networks, with knowledge of the technologies, and understanding of how those technologies may be
implemented.
Overview of WAN Environments

In the sphere of wide area networking there are two basic environments or domains; the public domain
and the private domain. Consideration for public or private, and surrounding security, must be evaluated
at all levels of networking. For example, most would consider the local connection between a private
facility, the customer site, and the carrier, to be private. But even this private line is carried in a public
domain to some extent. That means we need to consider, at all levels, the risks for security and
redundancy. Copper wire can be tapped; but copper is the most common local connection, or local
loop, media. If one were to use a wireless connection the problem is more readily seen. Consideration
does not necessarily mean implementation but does still indicate that a consultant perform a risk
assessment.
Using public carrier network infrastructure can be more cost effective than using privately owned
infrastructure, but this is all dependent upon the customers relationship with the carrier and what they
already have negotiated and may currently be using. In general, public carrier networks allow many
subscribers to share the costs of installing, managing, and maintaining the carrier infrastructure so that
often times they are lower in cost to each customer using that infrastructure.
Often times the two domains, public and private, are combined to gain the best of both. For example, a
customer may want to consider some redundancy between sites. This redundancy could take the form of
a primary private network that is backed up by the public Internet. The configuration could be such that
the private network is the primary route and the public Internet is secondary.
What is a WAN?
In the most general sense, a Wide Area Network (WAN) is a geographically dispersed telecommunications
network. For the purposes of this paper a WAN is generally defined as a network created to connect two
or more Local Area Networks (LANs). WAN discussion could include the interconnection between carriers,
but this is beyond the scope of this paper.
Subscriber A
Subscriber B
Public
Carriers
Central
Office (CO)
Public
Carriers
Central
Office (CO)
CO
Public Carrier
Network
Router A
CO
Router B
LAN A
LAN B
New York
London
Figure 1: General Carrier Supported WAN in N.A. Needs figure title
WAN connections can connect LANs located in the same city or around the world. A public carrier
network is commonly used to create WAN connections between LANs in different parts of the world. In
most regions it is the Public Telephone and Telegraph (PTT) companies, which serve Mexico, Europe,
Asia, South America, and other parts of the world.
Subscribers
Subscribers
Carrier
Interconnect
PTT
PTT
International Boundary
Figure 2: Basic Infrastructure in Most Regions
In Canada and the United States the public carrier networks include what are called Public Switched
Telephone Network (PSTN).
Subscribers
Local Exchange
Carriers (LECs)
CO
Interexchange Carriers
(IXCs)
LEC
Local Exchange
Carriers (LECs)
LEC
IXC
Subscribers
CO
IXC
Subscribers
Subscribers
CO
LEC
LEC
CO
Carrier Interconnects
Local Access and Transport

Area (LATA)
InterLATA
Local Access and Transport

Area (LATA)
Figure 3: Basic Infrastructure of North America
Over time, through the advent of the Internet and those companies providing services, the carrier who
previously may only have been connecting the lowest layers, have taken on the name of Service
Provider. Today the terms carrier and service provider or Internet Service Provider (ISP) are
sometimes used interchangeably. The ISP can provide basic local connectivity to them, and then further
connectivity into the Internet.
Over the past few years there has also been a blending of terms to the point that some technologies once
considered MAN (Metropolitan Area Network) are now included in some WAN discussions. This paper
focuses on the traditional and historical use of the acronym and therefore deals with the lower speed local
loop technologies (If one can rightly call ADSL2+ at 25Mbps a lower speed technology!) and does not
discuss SONET/SDH. This cross over, or blending, often occurs from the fact that the WAN needs a WAN
as well. The WAN of choice between carriers must of necessity be of much greater speed and capacity
since it serves as the core of all the customer networks. This is the realm of the telecommunications
providers and is beyond the scope of discussion for this paper.
How is LAN different from WAN?

Those familiar with LANs should not abandon all knowledge about them during their quest to understand
WAN design. Generally speaking a WAN differs from a LAN in areas regarding reoccurring costs (price),
performance, and span:
Price, since there is often a recurring cost to building a WAN. A LAN is typically installed and
the customer owns the wire and underlying switches. In a WAN you work with a vendor
(carrier or service provider) and pay them rent; the customer leases the lines and services
required to get from point A to point B.
Performance, since there are many differences at the physical layer, distances traveled, and
connection setup. LANs today are primarily Ethernet. There is no longer a major change
between Layer 1 on LAN such as from FDDI to Token Ring. LANs today are Ethernet. WANs
are not there yet. There are many flavors of Layer 2 used in the WAN. Therefore there will be
a need to convert Layer 2 and Layer 1, introducing latency. There will continue to be some
difference between the transmission speeds on a WAN also because WANs cover great
distance, and LAN which are often only 1000 meters from point A to point B. Even if a
service provider can provision some flavor of Ethernet, there will still be latency since the
distances are much greater. It is also impractical to use broadcast mechanisms for large
distances. Additionally WAN technologies must also handle connections, which often are
brought up only when needed, keeping costs lower but increasing the time for the first packet
to arrive at its destination.
Span, since WANs connect across vast distances that have no other end-points between point
A and point B and often cross oceans to bring point A and point B together. Span
means more than distance it means population density. A LAN has a dense population of endnodes on a LAN. A WAN is really a network of point to point, coterminous linkages, whether
physical or virtual.
Further consideration for the WAN environment is that data is not transmitted until there is a connection.
WAN connections are established at either layer 1 or layer 2 or both. In a LAN the remote end station is
always considered to be there. This led to the send and pray phrase that meant you could send data
but not always know if it got there. Obviously Layer 4 protocols such as TCP would accommodate this,
but others such as UDP would not. The main reason connections must be established for a WAN is
because no one wants to send data at great distance, pay for the travel charges, and then not have the
data be received. Imagine yourself attempting to travel from Singapore to Sydney, arranging the travel,
traveling, arriving, going to the final destination hotel, and it was shut down due to reconstruction. Quite
the expense with no results! At least you wouldnt be dropped by the destination! Of course there are
many permanent linkage options in WANs, but you will pay for the privilege of having the link always
connected.
Types of WAN Circuits

Subscriber A
Public Carrier
Network
Subscriber B
Dedicated
Physical
Circuit
Switched
Physical
Circuit
Permanent
Virtual Circuit
Switched
Virtual
Circuit
Figure 4: Types of WAN Circuits Illustrated
As the figure above shows, there are four types of circuits used in creating WAN connections when
considering both the physical and data link layers:
Dedicated physical circuits

Switched physical circuits
Permanent Virtual Circuits (PVCs)
Switched Virtual Circuits (SVCs)
Dedicated Physical Circuits

Dedicated circuits are permanent circuits dedicated to a single subscriber. The connection is always
active. The subscriber purchases dedicated time slots, or channels, that provide a specific amount of
bandwidth that is always available for the subscriber to use. The channels in a dedicated circuit are
created using time division multiplexing (TDM), which is discussed later in this section. In addition to
providing guaranteed bandwidth at all times, dedicated circuits provide the most secure and reliable WAN
connections available.
Switched Physical Circuits

Switched physical circuits are connected upon proper signaling exchange; for example, a phone call.
These circuits are switched on or connected between customers as the call routing demands. The
connection is active until one side or the other hangs up. (Next time you are troubleshooting a call
duration problem between routers you could just say the other router got angry with you and hung up!).
10
Seriously though, analog modems and ISDN circuits operate like this. The subscriber purchases the
ability to use the circuit but does not pay for the call duration time unless connected. This is what makes
these types of circuits useful for backup links.
Permanent Virtual Circuits (PVCs)

PVCs are also permanent circuits dedicated to a single subscriber. The connection is always active.
However, because multiple virtual circuits share a physical circuit, there is no guarantee that any specific
amount of bandwidth will be available at any specific time. Sometimes there may not be any bandwidth
available on the physical circuit because the physical circuit is saturated.
When the physical circuit is saturated, the traffic is temporarily stored at a switching point until
bandwidth becomes available. When bandwidth becomes available, the stored traffic is forwarded to its
destination. This process is referred to as store-and-forward processing, or packet switching, which is the
same processing method used on LANs.
PVCs provide an average bandwidth guarantee. The average bandwidth guarantee is accomplished
through statistical multiplexing (STM), which underlies packet switching technology. Because PVCs are
more cost effective for the public carrier, PVCs are usually less expensive for the subscriber than
dedicated circuits. PVCs are commonly used for Frame Relay, which is explained in detail in Frame Relay
section.
Switched Virtual Circuits (SVCs)

SVCs are identical to PVCs in all respects, except that they are temporary physical circuits. SVCs are
activated when a subscriber initiates a connection to transmit data. When all data have been transmitted,
the connection is deactivated, and the physical circuit resources are made available to other subscribers.
Because of these considerations the WAN is typically built up of many point to point connections, at both
layer 1 and layer 2. This can make it difficult for the designer to consider connectivity. To make the
routing most efficient the layer 2 network must often be fully meshed, to reduce the number of hops
between sites. (A full mesh is one where all sites are completely connected to every other site.) If all
traffic goes back and forth from central site to remote, there is little problem. When all sites have to
share information equally the number of interfaces required per site, physical or virtual, will be N1=interfaces, where N equals the number of sites.
N2
N1
N3
Figure 5: Simplified WAN Physical Layer Connection Paths
This is a simplified look at a private WAN physical layer possibility. Virtual circuits, such as created with
Frame Relay, will add another layer of complexity to this and will add connection points if you want a full
mesh as described with the physical layer example. Notice the example with Frame Relay. Even though
there is only one physical connection there are two arrow points at each physical connection. In essence
the same formula applies. It is just that you will need to consider both the raw physical bandwidth
available on a physical single link and then the committed information rate for the two virtual links.
11
N2
Layer 2 Frame Relay

Virtual Circuit paths.
N1
Frame Switch
N3
Figure 6: Frame Relay Connection Paths
How is LAN similar to WAN?

Generally speaking a WAN and LAN are similar when considering resource placement. One needs to
analyze the traffic flow regarding client to server communication. Some of these things can be estimated
(which is required for new installations), sometimes measurement tools should be put in place, but this is
not always possible as with a brand new installation. A good discussion with the customer will help
determine where the resources are and how often they are accessed by the clients. You will need to
understand the customers use of the following:
Network protocols and interconnection
architectures in general such as Bridging,

IP, TCP, and UDP.
How certain services and infrastructure
work such as their Web Server or email.
What are their requirements for security?
The topological layout of the WAN how

you want locations to interconnect.
The cost and performance offerings from
various service providers and/or carriers.

The logistics and planning for
deployment.
You should also consider how the customer network might change over the next months and years in an
attempt to allow for that in your planning. It is not possible to predict this with complete accuracy, but
will allow you to consider the options.
Note: This paper discusses the standard local loop technologies and the specified bandwidth available for these technologies. Although
physical layer bandwidth cannot be increased on an E1 or T1, or exceed certain limits with technologies such as ADSL, there are more
than these physical standards that are available to the designer. On the ProCurve Secure Routers 7102dl and 7203dl one can use other
link technologies and the 10Mbps serial interface module to connect to special modems. Some parts of the world offer wireless
transmissions that could run close to 10Mbps. Other parts of the world offer dark fiber to their modems to connect to the serial port.
Bandwidth can also be increased though the use of multilink protocols such as Multilink PPP or Multilink Frame Relay, more on those
later.
With all of this in mind lets stick with the focus of our paper first. Other considerations can be made
later. At this point lets consider performance, getting the data from point A to point B. One simple
illustration of this might be seen as the current bandwidth requirement at the central site is only 2Mbps
with four remotes feeding it at about 500Kbps each, but you have asked the right questions to determine
the four will double to eight in the next six months. You should plan on a minimum of 4Mbps at the
central site, and even that may be a little short sighted if you truly double every six months!
We could give you plenty of information in order to help you make the intelligent choice for the network
you are designing. The better solution, at least in our estimation, would be to couple the right amount of
information with a few examples. Ultimately you need to adapt these to your given situation.
12
Site A
Site C
User email downloaded
onto each end system.
Approximately 2 MB per
user system per day.
Email server
Accounting System
.
.
.
Site B
User email downloaded
onto each end system.
Approximately 2 MB per
user system per day.
Figure 7: Overly Simplified Sketch of Customer Requirements
A simple approach is to take the customer requirements and begin to sketch them out so that you can
see the big picture. In this overly simplified sketch we see that at minimum Site A and Site B will require
2 Mbytes of data in one direction per day per person. Downloading email over an eight hour period that
would become:
(2,000,000 bytes * 8 bits/byte) / 8 hours * 60 minutes/hour * 60 seconds/minute = 556 bits/second.
If you have 100 users you can see this is 55.6 Kbps. That is, as you will learn later in this paper, about 1
DS0 on an E1 or T1 channel.
But of course nothing is as simple as we design guide writers stipulate! So with this simple example we
have assumed that email is downloaded over the entire 8 hours evenly, but it never is. That is where you
come in and adjust up the required bandwidth for bursts and other traffic. It is more likely that between
8:30 and 8:45 AM that at least half of the email is downloaded for the entire day. All of a sudden our 8
hour window is compressed into 15 minutes and must handle 1 Mbyte per user at that time. When you
do that same equation you find:
(1,000,000 bytes * 8 bits/byte) / 15 minutes * 60 seconds/minute * 100 users = 889 K bits/second
Now the requirement is nearly a full megabit! But the rest of the day they need very little bandwidth.
One thing to note here, this is a simplified and hypothetical example. There are other factors that will
shave performance from the raw numbers. One should typically plan for at least a 30% hit simply from
packet headers and protocol handshaking. That now means that our requirement of 889 K bps is really a
full T1 of at least 1.544 Mbps. 1.544 Mbps less 30% is about 1 Mbps which if rounding is close enough
for this type of design to 889 Kbps.
Familiarity with the technologies will allow the consultant to help the customer make decisions about the
physical and data link layer requirements. Lets now look at those lower layer technologies.
13
Public Carrier
Network
Subscriber's
Site No. 1
WAN
Router
LAN
Local
Central
Office
Intermediate
Central Office
Demarcation
Point
Subscriber's
Site No. 2
Local
Central
Office
WAN
Router
LAN
Demarcation
Point
Figure 8: Basic Representation of WAN Path between Sites
At this point we should consider that there are always two sides to every story. Up to this point we have
not discussed that there are two sides to every conversation and that the two sites involved in the
conversation are subject to a myriad of considerations for performance between them. Some of the
things to observe and consider in the figure showing the basic representation of WAN path between sites
are:
What is the number of clients or servers
on site one that need access to site two?
What is the raw bit rate of the traffic flow

from site one to site two?
Is this transmission rate different when

sending from site one to two compared
to reception? Some technologies, such
as ADSL which well cover later, are
asymmetrical.
Is the full bit rate available for data or do

the sites under consideration allow for
some channels of the T1 or E1 link to be
used for telephony?
Is site one primarily comprised of a
single VPN client that requires access to

site two?
This final question is an interesting one to consider more closely. For example, is site one primarily
comprised of a single VPN client that requires access to site two? In this case the client VPN perspective
is just as important as the router side. The multitude of connection points between client and router in
this case is an ominous factor. Considering all points in the path between point A and point B are
essential in properly explaining the real performance potential of this path. What bandwidth is available
to the VPN client of site one?
Considering the single VPN client on a home or small business network it doesnt matter too much what
the link speed from the client to the router is, but the limiting factor would be the WAN link. That client
is going to get all the bandwidth to the router to use for their transmission. They share with only one
person so they get 100% of available LAN bandwidth. On the other hand, if there are many clients
funneling data through that router, then the link to the router is technically shared by all the users even
though it may be coming from a link on a switch! Regarding the aspect of sharing the data path we see
that WAN links are similar. The WAN link is shared by all others require data through it.
Another characteristic differentiating WAN from LAN is that the upper layer protocols, more specifically
the dynamic routing protocols, should be constructed to run efficiently. Responsiveness to routing
changes can come at the expense of delivering critical data traffic. Advances in Ethernet throughput in
LAN technology over the past few years has dramatically reduced the need to consider congestion, since
switched Ethernet congestion (which is no longer really Ethernet as we once knew) caused by running at
0.1, 1, and 10Gbps speeds is not typically congested at the physical and data link layers. We have grown
accustomed to so much speed that there was little need to consider congestion on a link! This is different
for WAN. Dynamic routing protocols can consume valuable bandwidth unless properly architected and
14
since the customer will always pay for renting or leasing of WAN links, they will always pay a price for
dynamic routing protocol overhead. Techniques such as route summarization and proper use of OSPF or
BGP can effectively control the potential problem.
Designing the Physical and Data Link Layers

An Overview of the Local Loop (The Transmission
Technologies)
Executive Summary
The local loop is the connection from the customer site to the service provider or carrier. They typically
work across 2 or 4 wire copper links, though fiber is also used. They are fixed bandwidth (T1, E1, ISDN,
and some types of DSL), or are variable bandwidth such as ADSL. Usable bandwidth for T1 is up to
1.536 Mbps, E1 is 1.92 Mbps, and ADSL up to 25Mbps downstream but is quite dependant upon distance
and link conditions and that is only in one direction.
Costs for these general transmission technologies vary globally. Without cost as a consideration the
choice will be for using guaranteed fixed bandwidth. These are typically T1 or E1 circuits. When cost is a
consideration, often ADSL or ISDN will be the link of choice. Some parts of the world can only get ISDN
so the question is answered by default for those areas.
Overview
All WAN connections consist of three basic elements:
The physical transmission media.

Electrical signaling specifications for
generating, transmitting, and receiving

signals through various transmission
media.
Data-linklayer protocols that provide
logical flow control for moving data

between peers in the WAN. (Peers are
the devices at either end of a WAN
connection.)
Note: This is a brief overview. If you need further information please see the references at the end of this paper.
The physical transmission media and electrical specifications are part of the physical layer (layer one) of
the Open Systems Interconnection (OSI) model, and data-linklayer protocols are part of the data-link
layer (layer two). They are used to create WAN connections into and through public carrier networks.
The connection between a subscribers premises and the public carriers nearest central office (CO) is
referred to as the local loop. The local loop includes the entire telecommunications infrastructuresuch as
repeaters, switches, cable, and connectorsrequired to connect a subscribers premises to the CO.
Public carrier networks were originally designed to carry analog voice calls. Therefore, copper wire is the
most common physical transmission media used on the local loop. Because of the limits in the signalcarrying capacity of copper wire, local loops that use copper wire are the slowest, least capable
component of a WAN connection. Public carriers are beginning to install coaxial and fiber optic cable in
local loops to meet ever-increasing bandwidth demands.
15
LAN
A Router (DTE)
Serial V.35
or X.21
Network
Interface Unit
(Smart Jack)
CSU/
DSU
B
Demarcation
Router with internal
Point
CSU/DSU
Public Carriers
CO
Repeater
OCU
Wire
Span
Wire
Span
Office Channel Unit

(Public Carriers CSU)
Figure 9: Infrastructure Common to Carrier Local Loops
T1 and E1 Technologies
T1 and E1 are basically defined as series of DC pulses that are generated at the rate of 1,544,000 pulses
per second for T1, or 2,048,000 pulses per second for E1. At the time of the introduction of T1 many
decades ago, it was designed to utilize the existing cable infrastructure in North America. The T1 bit
rate was determined due to the degradation of the signal across the two pair cables (TX-RX). The design
engineers determined that a 1.544 Mbps signal rate was the highest they could regenerate at 6000 ft.
(the distance between manhole covers in a typical U.S. city). These pulses could then be grouped
together and Channelized through Time Division Multiplexer techniques to carry 24 separate channels.
E1 differs from T1 primarily because it has 32 channels, a different encoding scheme, and different
framing.
Depending on the particular use of T1 or E1, you may see differing rates published. For example you
may only have 30 or 31 64Kbps channels available on an E1 circuit since some of the channels may be
used for signaling. An E1 bit rate of 2.048Mbps is still valid, but the bit rate is not equivalent to usable
bandwidth since encoding, framing and signaling overhead are factors detracting from the raw bit rate
available. More on this later.
A carrier WAN T1 or E1 connection provides a permanent, dedicated, point-to-point, fixed-bandwidth link
between two endpoints. Unless the service provider changes the path, the data sent between the two
endpoints in a carrier line WAN connection always flow along the same physical path.
The bandwidth for each connection is guaranteed across all parts of the path, because each connection is
allocated dedicated time slots, end-to-end. If there is no traffic to transmit, the time slots for that
connection go unused.
Note: T1 and T3 carrier lines are used primarily in Canada and the United States. In Europe and other Sector locations that follow the
ITU Telecommunications Standardization (ITU-T) standards, the comparable dedicated, high-speed WAN connections are E1 and E3
carrier lines. J1 and J3 carrier lines were defined for use in Japan.
T-carrier WAN connections are based on the American National Standards Institute (ANSI) T1.102 and
T.107 specifications. A T1 WAN connection provides twenty-four 64Kbps DS0 channels for a total of 1.544
Mbps as a data rate. After this point, dependant upon the framing and formatting chosen, the available
data rate can fall as low as 1.344Mbps (AMI encoding), or more typically today 1.536 (B8ZS encoding).
The loss of 8Kbps is formatting overhead. ProCurve Secure Routers default to B8ZS.
A full T1 connection uses all 24 DS0s. Fractional T1 connections, which use fewer than 24 DS0s, are also
available. The channels in a T1 connection can be used for voice traffic, data traffic, or a combination of
the two, but all traffic moving through the connection is in digital form.
16
Note: In North America, a subscribers site is connected to the central office (CO) of a local exchange carrier (LEC) that provides the T1
WAN connection. T1 WAN connections can also be created through multiple LECs and interexchange carriers (IXCs), as needed, to link
two subscribers premises together.
E-carrier lines are based on a range of specifications from the ITU from G.703 to G.822. An E1 WAN
connection provides thirty-two 64Kbps DS0 channels with 2.048 Mbps in total bit rate.
The bit rate of an E1 WAN connection is greater than that of a T1 WAN connection simply because there
are more DS0 channels (64K channels) available for data. There are differences in framing and encoding,
but collectively the E1 DS0s provide 32 times 64Kbps or 2.048Mbps. These 32 DS0s are collectively
referred to as bandwidth, yet the precise amount available as raw bandwidth is dependent upon the
usage for the circuit. An E1 effectively provides either a total of 1.984Mbps or 1.920Mbps, or 31 and 30
channels respectively, dependant upon whether the design requires use of channel 16 for signaling.
Some of this potential overhead comes from the use of channel 16 for signaling. Channel 16 signaling is
referred to as TS16 (Time Slot 16) signaling. So a full E1 carrier WAN connection uses either 30 or 31
DS0s. Fractional E1-carrier WAN connections, which use fewer than 30 channels, are also available. The
channels in an E1 WAN connection can be used for voice traffic, data traffic, or a combination of the two,
but all traffic moving through the connection is in digital form.
E1 is available in balanced (120 Ohm with BNC connectors) or in unbalanced mode (75 Ohm with RJ45
connector).
J-carrier WAN connections are a closely related variant of T-carrier WAN connections. i
Please use the reference material listed at the end of this paper for further information about T-carrier, Ecarrier, and J-carrier technologies.

Actual bit rate for T1 is 1.544 Mbps, E1 is
2.048 Mbps. Usable bit rate that can be
considered bandwidth is as low as
1.344 for T1 and 1.920 for E1 yet these
numbers vary with use model and
implementation.
Both are fixed bandwidth.

The links are always active by default a
permanent circuit.
How This Technology is Used

To connect from customer site routers to
carrier or ISP.
It can carry both data and traditional

voice conversations.
Multiple T1 or E1 links may be combined

to make one larger logical interface
through the use of layer 2 protocols like
Multi Link Frame Relay and Multi Link
PPP.
Advantages
The advantages to the T and E carrier technologies are because of fixed bandwidth, dependant upon the
technology. Bandwidth is constant and is available symmetrically. These technologies are well
established and interoperability should rarely be a concern once the proper encoding and framing are set
on both ends of the link.
Disadvantages
Speed relative to ADSL. See ADSL discussion following this section. ADSL can, under proper conditions,
deliver many times the performance of E1 or T1 but only asymmetrically.
What to Determine During Planning or for Implementation

Cost for given distance covered on each
link
Channels available for your data, all 24,

or 30, or some fraction?
Frame Type? For T1 the ProCurve
Secure Router 7000dl series supports D4

(SF) or ESF. For E1 the ProCurve Secure
Router 7000dl series supports FAS with
optional CRC-4.
17
Will voice services also be carried on this

link?
Encoding type? For T1 the ProCurve
The ProCurve Secure Router 7000dl

series currently supports the RJ45
connector for E1.
Secure Router 7000dl series supports

AMI or B8ZS. For E1 the ProCurve
Secure Router 7000dl series supports
AMI or HDB3.
ADSL Technology:
ADSL, and xDSL technologies in general, provide high-speed WAN connections over existing local loops.
To increase the amount of data that can be transmitted over the local loop (which is typically comprised
of plain copper wires), xDSL technologies employ advanced modulation techniques.
ADSL, in particular, was developed to alleviate a critical problem facing public carrierscongestion in the
public carrier network. With the increasing popularity of the Internet, more and more businesses and
residential customers began to connect to the Internet through the public carrier network. Because the
public carrier network was designed to handle random, short-term phone calls, carrying the traffic
created by numerous, lengthy Internet connections began to overwhelm the voice switches in the public
carrier network.
ADSL is only one of many types of DSL technology. Historically, as DSL technologies developed, the
collective group were often referred to as xDSL where the x is replaced with a letter that represents a
particular type of DSL, such as ADSL (Asymmetric DSL), HDSL (High bit rate DSL), and Very high bit rate
DSL (VDSL). The various types of xDSL provide different speeds, and the speed necessarily determines
how each type of xDSL is used. Over time the xDSL reference has changed and is simply now referred
to as DSL when discussing the collective group of technologies.
Because DSL works over existing local loops, it is a cost-effective WAN technology for both public carriers
and customers. By performing minimal adjustments to the existing copper lines that are used for most
local loops, public carriers can offer customers a high-speed broadband connection. In addition, DSL does
not require repeaters as T1 or E1, so it is less costly to implement than other traditional local loop
technologies. DSL is also an attractive solution for a wide range of customers, from residential customers
to large corporations.
With DSL the connection is always on. For customers who have used dial-up connections, this is a distinct
advantagesaving time because there is no dial-up process and eliminating the frustrations (such as
busy signals and disconnections) often associated with dial-up connections.
DSL has some disadvantages, however. For example, in the past, DSL has suffered from a lack of
standards, or better put, a lack of agreement on which standards to implement. Equipment was often
proprietary and did not interoperate. This is changing as standards groups further refine specifications for
various types of DSL.
In addition, DSL is not available in all areas because it is a distance-sensitive technology. If a company or
home is too far away from the public carriers central office (CO), DSL is not an option. The distance
between the company or home and the CO also dictates DSL transmission rates. The greater the
distance, the slower the rate.
DSL WAN connections can be either symmetric or asymmetric, depending on how data is transmitted
upstream and downstream. Downstream refers to the traffic being sent from the service provider or
public carrier to the customers premises. Upstream refers to the traffic being sent from the customers
premises to the service provider or public carrier.
If a DSL technology is symmetric, data is transmitted at the same speed both upstream and downstream.
This is sometimes called duplexed DSL. To avoid confusion with the more mainstream use of duplexing
(bidirectional transmissions), the term duplexed DSL is not used in this paper. Companies should select a
symmetric DSL solution for environments such as the following:
The DSL WAN connection is linking two office sites and equal amounts of data are transmitted
to each site.
18
Companies need to provide high-speed access to their network or web servers. In this case,
the upstream transmission speed would affect customers ability to access and download
information from the companies servers.
If a DSL technology is asymmetric, it provides different transmission speeds for upstream and
downstream. The transmission speed for downstream is higher than the transmission speed for
upstream. This makes asymmetric DSL ideal for Internet use because customers typically download more
data from the Internet then they upload. Below are tables of both asymmetrical and symmetrical DSL
technologies with their typical speeds, distances, and usages.
DSL Technology
Speed
Distance
Usage
IDSL
Up to 144 Kbps
5.49 km (18,000 ft.)
Internet access, video, telephony,

IP telephony
HDSL
1.544 Mbps (T1)

2.048 Mbps (E1)
2 pairs of wire; 3.66

4.57 km (12,000
15,000 ft.)
T1/E1 local loop, WAN connection

for businesses
HDSL2
1.544 Mbps (T1)

2.048 Mbps (E1)
2 pairs of wire; 3.66

4.57 km (12,000
15,000 ft.)

for businesses
SDSL
1.544 Mbps (T1)

2.048 Mbps (E1)
3.05 km (10,000 ft.)

for businesses
SHDSL
2.3 Mbps
1 pair of wire; 5.49

km (18,000 ft.)
WAN connection, video,

multimedia
VDSL*
Up to 34 Mbps
.3051.37 km
(1,0004,500 ft.)
Multimedia, HDT
* Can be either symmetric or asymmetric; usually asymmetric

Table 1: Table of Asymmetrical DSL Technologies
DSL Technology
Speed
Distance
Downstream: 1.5 to
8 Mbps
3.66 5.49 km
Upstream: Up to
1.544 Mbps
(12,00018,000 ft.)
Downstream: 1
Mbps
5.49 km (18,000 ft.)
Internet access, remote, LAN

access, VPNs, VOIP
ADSL
ADSL Lite
(G.Lite)
Usage
Internet access, video telephony,

IP telephony
Upstream: 512 Kbps

Downstream: 1.5 to
8 Mbps
3.66 5.49 km
Upstream: Up to
1.544 Mbps
(12,00018,000 ft.)
Internet access, remote, LAN

access, VPNs, VOIP
RADSL
19
DSL Technology
Speed
Distance
Downstream: 12
Mbps
3.84 5.67 km
Upstream: Up to
1.544 Mbps
(12,60018,600 ft.)
Downstream: Up to
25 Mbps
1.52 km (5,000 ft.)
Usage
Internet access, video, remote LAN

access, VPNs
ADSL2
Internet access, video, remote LAN

access, VPNs
ADSL2+
Upstream: Up to
1.544 Mbps
Downstream: 13
52 Mbps
.305 1.37 km
VDSL*
Multimedia, HDTV
Upstream: 1.5 2.3
Mbps
(1,000 4,500 ft.)
* Can be either symmetric or asymmetric; usually asymmetric

Table 2: Table of Symmetrical DSL Technologies
ADSL is arguably the most standardized type of DSL available. ADSL also supports analog voice on the
local loop. This gives ADSL a clear advantage over DSL technologies because customers do not need a
separate pair of wires to transmit analog voice. Their existing telephone equipment can continue to send
voice traffic over the same pair of wires that carry ADSL traffic. In the ADSL standards, support for
analog voice is called ADSL over Plain Old Telephone Service (POTS), or ADSL Annex A.
In addition to supporting analog voice, ADSL supports ISDN traffic. Customers who have ISDN
equipment such as telephones and fax machines can continue using this equipment while moving their
Internet or WAN connection to ADSL. Support for ISDN is called ADSL over ISDN, or ADSL Annex B, and
is common in countries such as Germany where ISDN is widely implemented.
Regional
Broadband
Network
Public
Carrier
Network
Figure 10: Typical Infrastructure of ADSL WAN
This figure illustrates a companys ADSL WAN connection. The WAN router functions as an ADSL
transceiver, performing the modulation required to send data at ADSL speeds across the local loop to the
public carriers CO. At the CO, the DSLAM (Digital Subscriber Line Access Multiplexer) aggregates ADSL
20
connections from multiple customers and creates one high-capacity connection to the regional broadband
network. This regional broadband network provides the backbone to connect DSLAMs from multiple public
carriers and connects each DSLAM to the Internet.
Because ADSL supports analog voice or ISDN traffic, the local loop is a shared medium. In an ADSL
Annex A environment, telephones send analog voice over the local loop, and the WAN router sends digital
data. At the CO, the analog voice must be transmitted to the voice switch and then routed over the public
carrier network. The digital data, on the other hand, must be transmitted to the DSLAM and then routed
over the regional broadband network. At the customers premises, the analog voice must be sent to the
telephones, and the digital data must be sent to the WAN router.
To separate the analog voice from the ADSL data, a POTS splitter is installed at both the customers
premises and the public carriers CO. The POTS splitter filters the traffic at both ends of the local loop and
ensures that the analog voice and the ADSL traffic are sent to the appropriate device at each location.
In an ADSL Annex B or Annex C environment, ISDN equipment and the WAN router transmit data over
the local loop. At the CO, the ISDN traffic must be transmitted to the ISDN switch and then routed over
the public carrier network. The ADSL data must be transmitted to the DSLAM and then routed over the
regional broadband network. At the customers premises, the ISDN data must be sent to the ISDN
equipment, and the ADSL data must be sent to the WAN router.
To separate the ISDN data from the ADSL data, an ISDN splitter is installed at both the customers
premises and the CO. This splitter ensures that each type of traffic is transmitted to the appropriate
device at each location.
Figure 11: ADSL Internet Connection
As mentioned earlier, ADSL is ideal for Internet access. To enable this Internet access, the regional
broadband network must be connected to the Internet. In this figure, the DSLAM connects directly to a
broadband switch, which is connected directly to a broadband access server. The broadband access
server then connects directly to a core Internet router. As the name suggests, the broadband access
server authenticates customers accessing the Internet through the broadband access network.
This figure shows one possible way to connect the DSLAM to the Internet. The exact configuration varies,
depending on factors such as the following:
The capabilities provided by the DSLAM

The broadband network equipment that
The technology used to create the

broadband network
the public carrier owns
21
In addition to aggregating multiple DSL connections, new DSLAMs provide advanced capabilities such as
ATM switching. In this case, the DSLAM may be connected directly to the broadband access server or
even to a core Internet router. The DSLAM may also be connected directly to the core Internet switch if
the public carrier owns that switch.
Finally, the public carrier must configure the DSLAM to support the technology used to create the
broadband network. Because DSL was originally developed for use with ATM-based broadband networks,
this is still the most common architecture. In fact, when ADSL Lite is implemented without splitters, ATM
is required: ATM cells must be included within the ADSL Lite frames.
Despite this ATM legacy, some public carriers and DSL vendors are investigating and implementing other
technologies for the broadband network. For example, the broadband network could be an Ethernetaggregation network linked together by a group of high-capacity switches.
Proponents of Ethernet-aggregation networks point to benefits such as lower costs, enhanced scalability,
enhanced support for services such as multimedia, more quality of service (QoS) features, and greater
resilience. Public carriers in Asia have already begun implementing Ethernet-aggregation networks.
Even if a majority of public carriers begin to migrate their broadband networks to Ethernet-aggregation
networks, ATM will have an ongoing role in DSL networks for some time. There is a large installed base of
ATM-based broadband networks, and because DSL was designed to work with ATM, ATM protocols are
often exchanged between the DSL transceiver and the DSLAM.
Finally there are many different Annex specifications for DSL technologies. Two of primary importance
for ADSL are Annex A which is ADSL over POTS and the other, Annex B, which is ADSL over ISDN. Below
is a simple table comparison of the two supported standards:
Annex A ADSL over POTS
Annex B ADSL over ISDN
Connector RJ-11C
Connector RJ-11C (some countries use

an RJ-45 connector. Germany is one
example)
ADSL2 - ITU G992.3
G.DMT ITU G992.1
ADSL2+ - ITU G992.5
Multi-Mode Auto detect mode
G.DMT - ITU G992.1

G.LITE - ITU G992.2
Multi-Mode - Auto detect mode
READSL2 - ITU G992.3 Annex L
ATM Multiple Protocol over AAL5
(RFC2684)
ATM Multiple Protocol over AAL5

(RFC2684)
ATM Forum UNI 3.1/4.0 PVC
ATM Forum UNI 3.1/4.0 PVC
ATM Class of Service (UBR)
ATM Class of Service (UBR)
PPP over ATM (RFC2364)
PPP over ATM (RFC2364)
PPP over Ethernet (RFC2516)
PPP over Ethernet (RFC2516)
ATM F5 OAM
ATM F5 OAM
Table 3: ADSL Annex A and Annex B Comparison
Summary of Major Points for ADSL

Speeds for different types of ADSL
Asymmetrical bandwidth.
Always on.
Different Annex for ISDN.
22

To connect from customer site routers to ISP.
It can carry data while allowing traditional voice conversations on existing voice equipment.
Advantages
Because ADSL works over existing local loops, it is a cost-effective WAN technology for both public
carriers and customers. By performing minimal adjustments to the existing copper lines that are used for
most local loops, public carriers can offer customers a high-speed broadband connection. In addition,
ADSL does not require repeaters, so it is less costly to implement than other local loop technologies.
ADSL is also an attractive solution for a wide range of customers, from residential customers to large
corporations.
Customers, on the other hand, get a high-speed connection at a relatively low cost. For example, ADSL is
less costly than T1- or E1-carrier lines.
With ADSL, the connection is always on. For customers who have used dial-up connections, this is a
distinct advantagesaving time because there is no dial-up process and eliminating the frustrations (such
as busy signals and disconnections) often associated with dial-up connections.
Disadvantages
ADSL is not available in all areas because it is a distance-sensitive technology. If a company or home is
too far away from the public carriers central office (CO), ADSL is not an option.
The distance between the company or home and the CO also dictates ADSL transmission rates; the
greater the distance, the slower the rate. This makes if very difficult to plan for bandwidth.
This author has not heard of committed information rates with ADSL and Internet connectivity. There is
no guarantee of bandwidth.
What to Determine During Planning

How far are my sites from the carriers central office?
Do the local carriers supply ADSL to that location?
Can the design allow for asymmetrical bandwidth? Is most of the traffic flow in one direction?
Will the customer site connect through to the carriers IP packet network to be routed over the
Internet or into an ATM network for a private WAN?
The ProCurve Secure Router 7000dl series currently support ADSL2+ in both Annex A and Annex B.
Current Support for Annex A
Current Support for Annex B
ITU G.992.1 Annex A (G.dmt)
ITU G.992.1 Annex B (G.dmt)
ITU G.992.2 Annex A (G.lite)
ITU G.992.3 Annex B ADSL2 (G.dmt.bis)
ITU G.992.3 Annex A ADSL2

(G.dmt.bis)
ITU G.992.5 Annex B ADSL2+
ITU G.992.3 Annex L READSL2

ITU G.992.5 Annex A ADSL2+
ANSI T1.413 Issue 2
Table 4: Current ProCurve Support Capabilities
ISDN
ISDN is a dial-up or, switched circuit, technology for WAN connections that was originally intended to
support voice, data, fax, and video services over standard telephone lines. Although ISDN is a
23
multipurpose solution, its core strength today is the ability to dial for connection before data
transmission. This relatively high-speed dial capability makes it suitable for backup link scenarios.
In North America, ISDN appears to have a dwindling role as a primary WAN connection. Other parts of
the world use it more frequently. Many public carriers are promoting Digital Subscriber Line (DSL)
connections, rather than ISDN. There are at least two reasons for this trend: First, DSL transmits data
faster than ISDN does. Second, DSL does not overload the switches that handle voice traffic through the
public carrier network. Instead, public carriers use data switches and routers to transmit DSL data. For
more information about DSL please see the section above or the references at the back of this paper.
However, there is at least one region where ISDN is still frequently used as a primary WAN connection. In
Europe, many public carriers actively sell ISDN as a primary WAN connection. Because these public
carriers have replaced their analog switches with digital switches, they have the capacity to provide
ISDN.
Note: In most regions, however, companies are implementing ISDN as a cost-effective backup to a carrier line WAN connection. If the
carrier line WAN connection is unavailable, the WAN router can use the ISDN WAN connection to send data.
In addition to these traditional implementations, some public carriers are offering a special ISDN implementation for retail business that
need to get approval on customers credit cards. This special implementation is discussed in more depth in later in this module.
ISDN provides an end-to-end digital connection between the source device and the destination device.
Because ISDN is a digital connection, it is not limited to the 56 Kbps maximum dial-up speed of an analog
connection. Instead, ISDN provides transmission speeds of 64 Kbps and above. The exact transmission
speed depends on the type of ISDN service and the region in which the service is delivered.
Public carriers offer two ISDN services:
Basic Rate Interface (BRI)

Primary Rate Interface (PRI)
BRI ISDN provides a transmission rate of 64 Kbps or 128 Kbps, while PRI ISDN provides a transmission
rate of 1.544 Mbps or 2.048 Mbps. (The next sections describe these services in more depth.) BRI ISDN
is provided across the twisted-pair cable that is used for ordinary telephones. PRI is provided as a T1
connection in North America and Japan, or as an E1 connection in Europe and Asia.
On the local loop, ISDN requires at least Category-3 (CAT-3) unshielded twisted pair (UTP). The number
of wires required depends on the ISDN service that you purchase: BRI ISDN requires two wires, or one
twisted pair. PRI ISDN requires four wires, or two twisted pairs.
When ISDN is implemented, the local loop is set up for BRI or PRI service. At the public carriers central office (CO), the office channel
unit (OCU) multiplexes and de-multiplexes channels on the twisted pair wiring of the local loop. Like the channels for carrier lines, ISDN
channels are based on DS0 or E0 and created through time division multiplexing (TDM). With BRI ISDN, the OCU multiplexes three
channels. With PRI ISDN, the OCU multiplexes 24 or 32 channels, depending on the region.
Because ISDN is a dial-up connection, it establishes a switched virtual circuit (SVC) when the subscriber
initiates or receives a call. For the duration of the call, the physical path through the public carrier
network is fixed. However, when the call is terminated and a new call is made, ISDN establishes another
physical path through the public carrier network.
A separate signaling channel is used called D-channel to setup and release a data channel (B-channel).
The network layer of the D-channel has not been defined in such details as the lower layers, therefore
different protocol implementations exist which are sometimes referred to as switch types.
24
ISDN Equipment at the Subscribers Premises
Terminal
Equipment
Network
Termination
2
(NT2)
Network
Termination
1
(NT1)
Network
Interface Unit
(Smart Jack)
Terminal
Adapter
Public Carriers
CO
Repeater
OCU
Wire
Span
Wire
Span
U Interface
Office Channel Unit

S Interface
R Interface
T Interface
Demarcation
Point
Figure 12: ISDN Equipment at the Subscribers Premises
The equipment required on the subscribers side of the loop varies, depending on the region and the
public carrier that is providing the ISDN service. This section explains the equipment that is generally
used in an ISDN network.
Network Termination 1
On the subscribers side of the local loop, the Network Termination 1 (NT1) provides the physical and
electrical termination for the ISDN line. The NT1 monitors the line, maintains timing, and provides power
to the ISDN line.
In Europe and Asia, the public carriers supply the NT1 device. In North America, however, the subscriber
provides the NT1 device. Many vendors are now building the NT1 directly into ISDN equipment such as
routers.
PRI ISDN also requires a Network Termination 2 (NT2) device. NT2 provides switching functions and data
concentration for managing traffic across the multiple B-channels.
In many regions, NT1 and NT2 are combined into a single device. In ISDN terminology, the device that
combines these functions is called an NT12 (NT-one-two) or just NT.
Terminal Equipment
Any devicesuch as a telephone, fax machine, or routerthat connects to an ISDN line is called terminal
equipment. Two types of terminal equipment are associated with an ISDN connection:
Terminal equipment 1 (TE1)

Terminal equipment 2 (TE2)
TE1 devices are ISDN ready and can be connected directly to the NT1 or the NT2. TE1 devices include
routers, digital phones, and digital fax machines.
TE2 devices do not natively support ISDN and cannot connect directly to an ISDN network. TE2 devices
require a terminal adapter (TA) to convert the analog signals produced by the TE2 device into digital
signals that can be transmitted over an ISDN connection. TE2 devices include analog telephones and
analog fax machines.
25
S Interface
T Interface
U Interface
Network
Termination
2
(NT2)
Terminal
Equipment
Four Wires
Network
Termination
1
(NT1)
Four Wires
Two Wires
Figure 13: ISDN Interfaces
Equipment can be at any of the four interface points on the subscribers side of an ISDN WAN connection:
U interface
S interface
T interface
R interface
These interfaces define the mechanical connectors, the electrical signals, and the protocols used for
connections between the ISDN equipment.
U Interface
The U interface provides the connection between the local loop and NT1. For BRI ISDN, the U interface is
one twisted pair. For PRI ISDN, the U interface is two twisted pairs.
Because public carriers in Europe and Asia provide the NT1, these regions do not use the U interface. In
regions that support the U interface, there can be only one U interface on the ISDN network.
T Interface
The T interface is used to connect the NT1 to the NT2. This interface is a four-wire connection, or two
twisted pair. Each pair handles the traffic sent in one direction (see the figure above).
In the United States and Canada, the T interfacealong with the NT1 and NT2is frequently built into a
circuit board in an ISDN device such as a router. In other regions, the T interface is the first interface at
the subscribers premises.
S Interface
The S interface is used to connect the NT2 to the TE1 or TA. This interface is also a four-wire connection,
or two twisted pair.
On a BRI ISDN, the S interface is mostly implemented as a passive bus, allowing you to connect multiple
TEs and TAs to the ISDN WAN connection. If you use a passive bus configuration, that bus is a shared
medium. The TEs or TAs connected to the passive bus must take turns transmitting, and they must be
able to detect collisions. PRI ISDN does not support multiple TEs at the S interface.
The S and T interfaces are often combined as the S/T interface.
R Interface
The R interface is used to connect TE2 to the TA. Because there are no standards for the R interface, the
vendor providing the TA determines how the TA connects and interacts with the TE2.
Connectors
The public carrier typically installs an RJ-45 jack to connect the subscribers premises to the local loop.
ISDN supports RJ-11 connectors, but an RJ-45 connector is recommended.
The following lists the advantages and disadvantages of ISDN:
26

The rest of the world uses an S/T
Switched circuit technology

Data or, B channels, each use a DS0
North America uses a U interface
interface
A BRI is two B channels of 64K each.

A PRI is same bit rates as T1 and E1
To back up IP routed network primary links.
Advantages
ISDN takes advantage of existing copper
wiring, and setup requirements are not
extensive or expensive.
ISDN can be used for both voice and
Because ISDN is a dial-up service, you

do not pay for idle connection time.
Paying only when the line is in use is
beneficial for infrequent calls.
data transmissions, and bandwidth can

be regulated according to your needs.
Disadvantages
Because you pay for ISDN when the line is in use, it can be costly if the connection is
maintained for long periods of time
There is a lack of interoperability between ISDN devices from different vendors.

Options
Switch Type
Euro-ISDN (sometimes called NET3 or DSS1), AT&T 5ESS, Northern

Telecom DMS-100, or National ISDN-1
Channel Usage
Voice, data, or both voice and data
Line Number 1 (or sometimes more

specifically for ISDN it is called MSN,
Multiple Subscriber Number)
Telephone number ex: 64 000 00000
Line Number 2 (or sometimes more

specifically for ISDN it is called MSN,
Multiple Subscriber Number)
Telephone number ex: 64 111 11111
SPID 1 (North America Only)
Telephone number + numerals ex: 64 000 00000 11
SPID 2 (North America Only)
Telephone number + numerals ex: 64 111 11111 00
TEI
Assigned by public carrier and can be either dynamically or statically

configured
Table 5: Table of Information Needed for ISDN
Data Link Layer Protocols in the WAN (The Transport

Technologies)
Executive Summary
The Data Link protocols are used to ensure reliable data transmission of variable length packages called
frames. The frame is the first basic structure required to transport the upper layer protocols such as IP.
In the WAN routing process the frame and lower physical layer are exchanged at each network
interface. This is the essence of pure routing. WAN routers must not only reconstruct the lower layers to
27
forward layer 3 protocol but may have to re-establish connection before forwarding such as with any
switched circuit at either layer 1 or 2.
PPP and Frame Relay can also take advantage of link aggregation where as HDLC cannot. Link
aggregation is the combining of two or more physical links (interfaces) into one larger logical interface.
Overview
As mentioned previously all WAN connections consist of three basic elements:
1.
The physical transmission media
2.
Electrical signaling specifications for generating, transmitting, and receiving signals

through various transmission media
3.
Data-linklayer protocols that provide logical flow control for moving data between
peers in the WAN
For each of the local loop connections, a customer can choose among several data link layer protocols
such as HDLC, PPP, Frame Relay, or ATM; these are defined later in this paper. The Data Link protocols
may span only the local loop, span across regions, or even go intercontinental. This is unlike the physical
layer transmission technologies that are only concerned with moving electrical signals from customer
location to the central office for processing.
Data link layer protocols are chosen for a variety of technical and business reasons. Typically they are
chosen for the technology that allows the best communication between point A and point B
expediently for the given requirement. PPP (Point to Point Protocol), HDLC (High Level Data Link
Control), ATM (Asynchronous Transport Method), and Frame Relay are among the most common and
interoperability is typically not an issue.
If the customer leases their lines between sites so that the carrier is not involved in Data Link layer
switching or routing (other than simply allowing it to traverse the wires on their physical infrastructure),
and they are building a private non-Internet based WAN, then PPP or HDLC is often chosen. If a more
complex private WAN spanning many carriers is required then Frame Relay will likely be chosen.
If HDLC or PPP can be used there are two primary considerations: what the router on the other end of the
link supports and if there is a requirement for security (particularly for authentication of the connection
router). There are other minor factors but these are most prominent.
Some routers perform HDLC by default and the customer may not have required some basic security. In
this situation it is acceptable to choose HDLC. If you are connecting site A and site B and the router
at site B supports PPP, or PPP is already enabled, then choose PPP at site A. Often PPP is chosen to
allow for authentication of link establishment. So PPP allows for security and HDLC does not.
PPP and Frame Relay can also take advantage of link aggregation while HDLC cannot. Link aggregation
is the combining of two or more physical links (interfaces) into one larger logical interface.
HDLC
High-Level Data Link Control, also know as HDLC, is a bit oriented protocol. HDLC is one of the oldest
data link layer protocols for the WAN. In fact, it predates the PC and was originally developed for
mainframe environments. Because of this, HDLC was originally designed for use with primary and
secondary devices, such as a mainframe with dumb terminals.
HDLC is a protocol developed by the International Organization for Standardization (ISO). It is used
throughout the world and widely implemented because it supports both half duplex and full duplex
communication lines, point to point, and multi-point networks, on both switched or non-switched links.
HDLC is designed to permit synchronous, protocol transparent data transmission. HDLC also has many off
shoots. Some of these are Synchronous Data Link Control (SDLC) and Link Access Procedure-Balanced
(LAP-B), and PPP.

A point to point protocol
No security
28

Site to site transport of upper layer network.
Connection to Cisco routers.
Advantages
HDLC is a simple, effective, and well used standard for layer 2 point to point connections. Many, if not
all, WAN devices should support this protocol.
Disadvantages
It does not support the authentication and many other features of PPP. There is no capability in HDLC or
PPP to cross carrier boundaries. There is no NNI (network-to-network interface). These are advantages
that are covered in the following technologies but unfortunately need mentioned here without previous
reference.

HDLC in the ProCurve Secure Router 7000dl series currently allow for the links to be un-
numbered. Un-numbered is a capability that allows for one of the link IP addresses to use the
address of the Ethernet port. This special capability is useful to those conserving address
space. In design determine if the customer needs to have the link run in un-numbered mode.
Will this HDLC network require backup?
PPP
Point-to-Point Protocol (PPP) is the name of a single protocol, but most often PPP refers to the entire
suite of protocols that are related to PPP. PPP is a layered protocol, starting with a Link Control Protocol
(LCP) for link establishment, configuration and testing. Once the LCP is initialized, one or many of several
Network Control Protocols (NCPs) can be used to transport traffic for a particular protocol suite. The IP
Control Protocol (IPCP), documented in RFC 1332, permits the transport of IP packets over a PPP link.
PPP differs from HDLC primarily in that you can use some basic security methods with it. You can
configure WAN routers (or other devices) to use optional protocols in the PPP suite. In addition, many
protocols in the PPP suite, such as LCP, allow you to manually configure options.
When one of the peers in a PPP session has been configured to use protocols or options that are not used
by default, the peers negotiate these options. They do so by exchanging configuration frames for the
protocol in question.

Point-to-Point. One point to a single other point.
Some security available through authentication.
Site to site, point to point, transport of upper layer protocol, IP for the ProCurve Secure Router
7000dl series, over a leased physical link.
To establish an authenticated connection from router to ISP router over the physical connection
when often this connection is ATM over ADSL, or Ethernet through an ADSL modem.

The authentication method desired, typically CHAP or PAP.
The passwords used during authentication.
Advantages
PPP allows for a wide range of features one of which is authentication of the other network device
attempting to connect. This particular feature makes it extremely suitable for ADSL and Internet
connectivity.
29
Disadvantages
There is no capability in HDLC or PPP to cross carrier boundaries. There is no network-to-network
interface.

The PPP attributes for the router at the other end of the connection.
Determine if passwords and authentication methods are required (typically these are not
needed in a private physical network).
Frame Relay
Frame Relay is a high-performance, packet-switching WAN technology. Frame Relay carrier networks use
statistical multiplexing to allow multiple virtual connections to be multiplexed across a single physical link.
This allows the Frame Relay network to better manage and use the available bit rate of the physical layer.
Frame Relay can carry multimedia traffic, such as voice or video, but it is not an ideal WAN technology for
time sensitive applications. Frame Relay was designed for delay-tolerant traffic that may be bursty in
nature. Additionally the carriers usually only offer a limited Quality of Service. So a customer with only
data traffic is more likely to be satisfied when subscribing to Frame Relay than a customer requiring voice
and/or video transport coupled with their data.
Although Layer 2 protocols are primarily concerned with reliable data transport, Frame Relay was
designed when Layer 1 physical transmissions were growing in reliability. Frame Relay left out many of
the unnecessary error correction (retransmission of data) up to the end-points, which speeds up overall
data transmission. When an error is detected in a frame, it is simply dropped. The end stations are
responsible for detecting and retransmitting dropped frames. Frame relay allows data transmission in
variable sized packages called frames. This is different than any fixed length size such as used in ATM
or SONET/SDH.
With Frame Relay, a router at the customers site is configured to communicate only with another router
in a point-to-point connection. This connection is called a Virtual Circuit (VC). The Frame Relay standard
allows for either a Permanent Virtual Circuit (PVC) or a Switched Virtual Circuit (SVC), but most
customers only use PVCs. PVCs are established by using a carrier physical WAN connection or another
form of permanent connection. LMI (Local Management Interface) circuit status messages provide
communication and synchronization between Frame Relay DTE and DCE devices. These messages are
used to periodically report on the status of PVC. There are different implementations of LMI.
Like carrier lines, Frame Relay is most frequently implemented through a public carrier network. The
customer purchases a connection (E1 or T1) to a public carrier that provides Frame Relay service. The
public carrier owns all of the Frame Relay equipment and the WAN infrastructure in the Frame Relay
cloud.
When the customer purchases Frame Relay service, they negotiate a Service Level Agreement (SLA) that
specifies the amount of bandwidth they will receive. This bandwidth is called a Committed Information
Rate (CIR). The CIR can be considered a guaranteed amount of bandwidth. It is, however, contractually
guaranteed, rather than physically guaranteed as it is with dedicated T and E carrier lines that use TDM.
Frame Relay carriers usually rebate the customer if they do not provide the CIR commitment, so they are
generally careful to meet it.
Frame Relay has the added benefit of both user to network (UNI) and network to network (NNI)
interfaces. NNI allows transport between and through multiple carriers to go from point A to point B.
This does not affect the router configuration but allows the customer to understand what technology to
use. PPP and HDLC are limited in the sense that they do not have termination within the carriers at layer
2 and do not have an NNI. If a customer wants to use HDLC or PPP spanning multiple carriers, then they
must lease the physical connection between point A and point B.
Frame Relay also has the ability to perform multilinking with the implementation agreement referred to
as FRF.16. Multilink Frame Relay (MFR) allows the customer the ability to increase bandwidth by
enabling multiple serial links to be aggregated, or trunked, to form a larger logical link of greater
bandwidth. MFR allows for UNI and NNI Frame Relay networks.
30
The logical links formed are called bundles. Routers using MFR and their peer routers exchange link
integrity control messages to determine which bundle links are active and to synchronize those bundle
links that should be associated with each given bundle.
The bundle link interface's line protocol status is considered active and up when the peer router
acknowledges that it will use the same link for the bundle. The line protocol remains up when the peer
device acknowledges the hello messages from the local router.
MFR provides load balancing across the bundle links within a bundle. If a bundle link chosen for
transmission happens to be busy transmitting a long packet, the load balancing mechanism can try
another link, thus solving the problems seen when delay-sensitive packets have to wait.
MFR allows you to design a Frame Relay interface with more bandwidth than is available from any single
physical interface. For example, many new network applications require more bandwidth than is
available on a E1 or T1 line. Combining multiple E1 lines may be less costly than ordering a single E3 or
T3.
MFR can also provide for high availability through the use of multiple physical interfaces, but is primarily
a method for greater capacity.

Packet-switching technology
Designed for delay-tolerant traffic that
may be bursty in nature.
Erred frames are dropped and the end
stations are responsible for detecting and

retransmitting dropped frames.
Frame Relay uses Virtual Circuits (VC).
PVCs are established by using a carrier

physical WAN connection or another form
of permanent connection.
Statistically multiplexes many VCs onto a

single physical link.
SLAs are negotiated for bandwidth that is

specified as the Committed Information
Rate (CIR).
Has a Network to Network Interface,
NNI, that allows Frame Relay

communication across carrier networks.
MFR combines multiple physical links into

one logical link called a bundle.

Interconnection of multiple sites through a minimum number of physical links required at each
site.
Especially used when crossing carrier or national boundaries.

Advantages
Frame Relay packet switching enables Frame Relay carriers to allocate network resources to active
customers. Unlike Time Division Multiplexed (TDM) technologies, such as dedicated carrier lines and
ISDN, packet-switching technologies do not ensure a constant bandwidth. However, LAN and WAN traffic
is typically bursty, meaning large amounts of data are often transmitted at once, after which there may
be no traffic whatsoever for a period of time. Thus with TDM, when the network is not transmitting,
bandwidth is wasted; with packet-switching, when one network is not transmitting, another network can
use the bandwidth. As a result, Frame Relay is more cost-effective for Frame Relay carriers, making it
less expensive in many regions than dedicated carrier lines.
Frame Relay is also extremely flexible. Data in a Frame Relay network generally flows along the same
physical path. However, that physical path is controlled simply by software in a switch, rather than any
physical configurations. As a result, it is generally much easier to reroute traffic in the event of a switch
or line failure.
Disadvantages
Although sharing resources among many circuits in a Frame Relay network provides many advantages, it
also has one major drawback: Network throughput will be lower during peak hours if the Frame Relay
network becomes congested. The customer may consider increasing their contracted CIR or add extra
links.
31
PPP and other data-linklayer protocols, such as Frame Relay, establish a single point-to-point
connection, which may not provide sufficient bandwidth to meet a business requirements. Linkaggregation protocols such as Multilink Frame Relay address this limitation.
Theoretically, link aggregation is a simple idea: effectively double your available bandwidth by using two
physical cables to connect your endpoints instead of only one, triple your bandwidth by using three
cables, quadruple your bandwidth by using four cables, and so on. For example, you could aggregate two
1.544-Mbps T1 connections into a virtual single network connection with an underlying bandwidth of
3.088 Mbps.
However, to take advantage of multiple physical cables, data-linklayer protocols must be modified to
fragment frames into smaller frames that can be passed simultaneously over separate cables and then
reassembled by the receiving peer. Link-aggregation protocols, including Multilink PPP (MP) and Multilink
Frame Relay (MFR), do exactly that.

What is the minimum amount of
bandwidth needed at the site under

consideration? This yields the CIR?
What is the LMI Type
Does the customer need more bandwidth

than a single E1 or T1?
Does the carrier provide Multilink Frame

Relay in the area under investigation?
DLCI Numbers and how they are routed

to other sites
ATM Technology
Asynchronous Transfer Mode, ATM, is the most frequently used backbone technology in the world. It was
conceived as a multi service transport technology. It is a standards based transport technology that
is deployed broadly within the core of the Internet and carrier networks. It is also used frequently at the
edge of telecommunications systems to send data, voice and video.
ATM is best known for its integration with other technologies, and for its management features that allow
carriers to offer quality of service (QoS) guarantees. ATM has historically been referred to as cell relay
and uses short, fixed length, packages called cells for transporting information that is either data or
voice. This information is divided among these cells, transmitted and then reassembled at their final
destination. This cell based technology is designed to work over any physical layer technology. ATM
can be used in many applications. Customers can have flexible access to the network resources and can
achieve a reasonable concession between performance and cost.
ATM technology consists of many layers. The first layer, the adaptation layer, holds the bulk of the
transmission. This 48 byte payload divides the data into different types. The ATM layer contains five
bytes of additional information, referred to as overhead. This header information allows the ATM
switches to forward the cells to their destinations. Lastly, the physical layer attaches the electrical
elements and network interfaces.
Service Categories
The ATM service categories are defined within the ITU-T, and specified in detail by the ATM Forum. ATM
service categories make the technology suitable for an almost unlimited number of applications. ATM
provides Virtual Path (VP) or Virtual Channel (VC) Connections with different levels of service. Each
service level can be chosen to best meet the application requirement.
Most of the requirements that are specific to a given application may be resolved at the edges of an ATM
network by choosing an appropriate ATM Adaptation Layer or AAL. It should also be noted that although
the adaptation layer provides many services, they are not always required for all devices. Routers that
offer QoS features at the upper layers can provide similar functionality. (There are many ways to solve
any particular problem.)
The ATM layer behavior should not rely on the AAL protocols, since these are service specific, that is, the
upper layers should define and use the services they need. In many cases the services are supported by
the customer router. Given the presence of a heterogeneous traffic mix, and the need to adequately
32
control the allocation of network resources for each traffic component, a much greater degree of
flexibility, fairness and utilization of the network can be achieved by providing a selectable set of
capabilities within the ATM layer itself. The Service Categories have been defined with this goal in mind.
Definitions
The following definitions address certain issues regarding Traffic Management and Congestion Control and
are now stabilized standards managed by the ATM Forum. The ATM Forum forms a base to facilitate
interoperability in multi-vendor implementations of these standards.
A first classification of these services and capabilities may be seen from a network resource allocation
viewpoint here below:
Constant (maximum) bandwidth
allocation is called Constant Bit Rate

(CBR) in the ATM Forum and
Deterministic Bit Rate (DBR) by the ITUT.
Statistical (average) bandwidth allocation

is called Variable Bit Rate (VBR) by the
ATM Forum and Statistical Bit Rate (SBR)
by the ITU-T.
Further definition comes from ATM
Forum as it groups VBR into real-time

(rt-VBR) and non-real-time (nrt-VBR)
classifications depending on the QoS
requirements.
Another division actually defines three

VBR sub-classes depending on the
conformance criteria adopted.
Available Bit Rate (ABR) is used by both
the ATM Forum and ITU-T. It is an

elastic type of bandwidth allocation
that occurs when the amount of reserved
resources providing the bandwidth can
vary with time, governed by network
availability.
One category that is considered only in

the ATM Forum is the Unspecified Bit
Rate (UBR). No explicit resource
allocation is performed. There is no
bandwidth or QoS objectives specified.
One final category for consideration is
within the ITU-T only. It is based on

block (or burst) allocation and is called
ATM Block Transfer (ABT). This class
allows network resources to be
negotiated and allocated on a per block
basis rather than on a per connection
basis.
Note: this paper will use the ATM Forum category designations since they seem to be more familiar and widely used in publications.
Services
ATM services use parameters and procedures for controlling traffic and congestion. These services are
used to allow the network to achieve its performance goals. To meet these goals, services form the
framework for managing and controlling network traffic.
Service categories are used on a per connection basis and formed during set up of the connection.
Service categories work on both the Virtual Channel Connection, VCC, and the Virtual Path Connection,
VPC.
Traffic Parameters
There are certain traffic parameters that describe characteristics of any given router using ATM. They are
listed below for your reference and help with our understanding of Service Categories:
Sustainable Cell Rate (SCR)

Peak Cell Rate (PCR)
Minimum Cell Rate (MCR)
Maximum Burst Size (MBS)
And finally there are QoS Parameters
Note: that these parameters are reasonably self descriptive. For further information you should seek out references listed at the end of
this paper.
33
When QoS parameters are selected they relate to a given performance objective the customer has in the
network. These parameters are listed below:
Maximum Cell Transfer Delay (Max CTD)

Cell Delay Variation (CDV)
Cell Loss Ratio (CLR)
Cell Error Ratio (CER)
Severely Errored Cell Block Ratio (SECBR)
Cell Mis-insertion Rate (CMR)
Service Categories
Constant Bit Rate (CBR)
CBR is used to set a fixed amount of bandwidth. It is characterized by the PCR value and is always
available during the life of the connection. The router may send traffic at or below the PCR any time.
This category is intended for real time applications that require tight control over jitter.
Key traffic parameters for this category are PCR, CTD, CDV, and CES. When the connection is
established and the basic commitment is made by the network, the QoS negotiated is guaranteed to all
cells conforming to the conformance tests. It is assumed that cells which are delayed beyond the value
specified by CTD are less value to the application.
For example any data, text, or image transfer application that requires smooth traffic flow and response
time from the other side should have a CBR channel. Examples are:
Videoconferencing
Interactive Audio such as with phone conversation.
Real-Time Variable Bit Rate (rt-VBR)
Real-Time Variable Bit Rate, rt-VBR, is for time sensitive applications. Time sensitive applications are
those that require tight constraints on delay and variation in delay. Real-Time VBR serves voice and
video applications well, especially when voice or video sources are expected to have a transmission rates
that vary over time. Real-Time VBR would be useful to any application that could be characterized as
bursty.
Key traffic parameters for this category are PCR, SCR, and MBS. Cells that are delayed beyond the CTD
are assumed to be of lesser value to the given application. The rt-VBR service can also support statistical
multiplexing from real time sources.
Non-Real-Time Variable Bit Rate (nrt-VBR)
The non-real time VBR service category is for applications that may still have bursty characteristics
but do not have tight requirements for constraining variation in delay.
Key traffic parameters for this category are also PCR, SCR, and MBS. For the cells that are transferred
within the contractual agreement, the application can expect a low CLR. For all other cells there are
bounds on the CTD.
For example, VBR is suitable for any application for which the end system can benefit from statistical
multiplexing, by sending information at a variable rate, and can tolerate or recover from a potentially
small random loss of information.
Available Bit Rate (ABR)
ABR is a service category for sources that have the ability to increase or reduce their information rate
when the supporting ATM network requires them to. This allows the sources sending traffic into the ATM
network to take advantage of the changes in the ATM layer transfer characteristics (i.e., bandwidth
availability) subsequent to connection establishment.
There are many applications having imprecise requirements for throughput and can utilize ranges of
acceptable values; as with a minimum and maximum value rather than an average value. To do this an
ABR connection is used and the source specifies the minimum and maximum bandwidth requirement.
Key traffic parameters for this category are PCR, MCR, CLR, and CDV. Although no specific QoS
parameters are negotiated with ABR, it is expected that an end system adapts its traffic in accord with
34
the feedback it receives with a low CLR. CDV is not controlled in this service, although cells sent from the
source are not unnecessarily delayed. PCR and MCR are used to define the boundaries for the available
bandwidth.
For example, any non time critical application running on the end systems that are capable of varying
their emission rates can utilize the ABR service. This includes internetworking, such as one would have
with an IP router. Though ABR can be used for IP routing, another service category is often chosen and
described in the Unspecified Bit Rate.
Unspecified Bit Rate (UBR)
The Unspecified Bit Rate (UBR) category is best defined as delivering traffic on a best effort basis. It is
for servicing non critical applications. (Not that IP data is non critical but when you consider that often
TCP and IP are combined together this lower layer ATM service looks at it as non critical when
compared with other applications.) UBR, from the ATM perspective, is for those applications that do not
require tightly constrained delay and delay variation; or for that matter any particular level of QoS. UBR
can be used by IP routers when QoS is furnished at the upper layers.
There are no key traffic parameters for this category since UBR does not give any guarantees for traffic.
UBR provides an appropriate solution for less demanding applications. Most data applications such as file
transfers are highly tolerant to delay and cell loss.
Typical Application
CBR
rt-VBR
nrt-VBR
ABR
UBR
LAN Interconnection
PABX Circuit emulation
N/R
N/R
N/R
POTS or ISDN Video Conferencing
N/R
N/R
N/R
N/R
Compressed Audio
Interactive Multimedia
Table 6: Typical Applications and Ratings for ATM service categories

Note: Recommended usage for the typical application of a service category are: good=1, better=2, best=3. The service categories
that are not good for a given application are shown with a not recommended or N/R.
Obviously , many of these categories are outside the scope of this paper, that is focused on WAN design with IP routers, but are put
here for your reference and understanding.

ATM service categories can be used in different ways and in many combinations. They are used within
any VP or VC connection. These service help the customer gain the service level they require for the
given application. Most often with IP based networks the chosen service category is UBR because QoS
may be manageable with the upper layers.
ATM connections offer QoS at the lowest layers.

ATM works over any lower layer.
To connect from customer site routers to carrier or ISP.
It can carry both data and traditional voice conversations.
ATM is often combined with ADSL for IP routing.
Advantages
The real advantage for ATM is in its usage. For the customer requiring connection to the Internet, ADSL
and ATM combine to build the best solution. ATM also is specified to provide QoS at the lowest layers.
35
Disadvantages
There are no real disadvantages in connecting to an ATM infrastructure.

Since ATM and ADSL are often combined, many of these questions will be addressed below in a
separate section.
The PVC numbers that are available from the carrier.

Customer name or ID and password for your connection that will be used in the PPPoE
connection
Encapsulation for ATM, either AAL5mux or AAL5snap

Whether this is a PPPoE or PPPoA implementation
VPI / VCI used by provider for ATM over ADSL connectivity
Whether this is a PPPOE or PPPOA installation
Using ATM and DSL together

Probably the most used application with respect to IP branch office routing is the combination of ATM and
DSL together. Used together, these two technologies can take the best qualities of each and provide an
outstanding solution for customer interconnection to the Internet.
ATM is unique in its ability to work with DSL due to the number of services and applications it easily
works with. It allows different types of traffic to be transported over the same network while supplying
an unparalleled degree of QoS. DSL technologies use the ability of ATM to work over any lower layer
technology to deliver the services needed by the customer wanting to hook to the Internet. Many DSL
options are available with, as was discussed previously, ADSL being the most dominant currently.
ATM and ADSL are often combined. Below are some questions that should be addressed during planning
and before implementation:
PPPoA (Point to Point Protocol over ATM) Planning Questions

What is the modulation; DMT or CAP?
What are the PVC numbers VPI/VCI values?
For authentication with PPP, what is the username and password required for your router to connect
with?
If there is a static IP address assigned for your router, what is the public IP address and network mask?
If there is a static IP address assigned, what is the IP address of the providers default gateway?
PPPoE (Point to Point Protocol over Ethernet) Planning Questions

Will an end-station such as a PC or the router be the PPPoE client?
What is the domain name of your provider?
For authentication with PPP, what is the username and password required for your end-station or router
to connect with?
RFC 1483 Planning Questions

36
Further Information
For further information about layer 1, layer 2, and other networking protocols please see the references
at the end of this paper.
Solution Examples for Layer 1 and 2

PPP Solution Example
The customer has the need to connect to multiple sites. The local cost for T1 or E1 is reasonable at their
central location so multiple physical links are less expensive than other technologies. The remote sites
are within the boundaries of the local carrier. The customer decides to lease the physical links to both
sites. PPP is chosen at layer 2 but they could have used HDLC since there should not be any serious
security issues on private leased lines. The Cisco router, a 2621XM supports PPP so there are no
interoperability issues.
Note: Please note that these examples, and those in further sections, are given for your study and consideration only. They are to help
you reach a better understanding of the fundamental concepts before configuring your own application. It will be necessary for you to
modify these examples to match your own network design.
ProCurve 7203dl
Hostname: 7203_central
t1 1/2
10.10.2.252/24
PPP
Carrier
t1 1/1
10.10.1.254/24
PPP
Ethernet 0/1
192.168.3.253/24
ProCurve 7102dl
Hostname: 7102_right
t1 1/1
10.10.1.253/24
PPP
Cisco 2621
Hostname: 2621_left
Ethernet 0/1
192.168.1.253/24
ser 0/0
10.10.2.251/24
PPP
Figure 14: PPP Solution Example
ProCurve Secure Router 7102dl in PPP Network
!
!
hostname "7102_right"
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging forwarding
no logging email
logging email priority-level info
!
username "pnb" password "pnb"
!
!
interface eth 0/1
ip address 192.168.1.253
255.255.255.0
no shutdown
no lldp send system-description
37
lldp send management-address

!
interface eth 0/2
no ip address
shutdown
!
!
!
interface t1 1/1
tdm-group 1 timeslots 1-24 speed 64
no shutdown
!
interface t1 1/2
line-length 0
shutdown
!
interface ppp 1
ip address 10.10.1.253
255.255.255.0
no shutdown
bind 1 t1 1/1 1 ppp 1
!
!
router rip
redistribute connected
network 10.10.1.0 255.255.255.0
network 10.10.1.253 255.255.255.0
!
!
no ip tftp server
ip http server
ip http secure-server
ip snmp agent
no ip ftp agent
!
!
line con 0
login
password pnb
!
line telnet 0 4
login
password pnb
!
end
ProCurve Secure Router 7203dl in PPP Network
!
!
hostname "7203_central"
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
customername "pnb" password "pnb"
!
!
!
interface eth 0/1

ip address 192.168.3.253
255.255.255.0
no shutdown
!
interface eth 0/2
no ip address
shutdown
!
!
!
interface t1 1/1
no shutdown
38
!
interface t1 1/2
clock source through
no shutdown
!
!
interface ppp 1
255.255.255.0
no lldp send system-capabilities
no shutdown
bind 1 t1 1/1 1 ppp 1
!
interface ppp 2
255.255.255.0
no shutdown
bind 2 t1 1/2 2 ppp 2
!
!
router rip
network 10.10.1.254 255.255.255.0
network 10.10.2.241 255.255.255.0
!
!
no ip tftp server
ip http server
ip snmp agent
no ip ftp agent
!
!
line con 0
login local-customerlist
!
line telnet 0 4
password pnb
!
end
Cisco 2621xm in PPP Network
!
version 12.3
service timestamps debug datetime msec
service timestamps log datetime msec
no service password-encryption
!
hostname 2621_left
!
boot-start-marker
boot-end-marker
!
!
no network-clock-participate slot 1
no network-clock-participate wic 0
no aaa new-model
ip subnet-zero
ip cef
!
!
ip audit po max-events 100

no ftp-server write-enable
!
!
interface FastEthernet0/0
ip address 192.168.2.253
255.255.255.0
ip rip send version 1
ip rip receive version 1
duplex auto
speed auto
!
interface Serial0/0
ip address 10.10.2.251 255.255.255.0
encapsulation ppp
!
!
39
ip address 192.168.1.2 255.255.255.0
shutdown
duplex auto
speed auto
!
router rip
network 10.0.0.0
network 192.168.2.0
!
ip classless
ip http server
no ip http secure-server
!
!
control-plane
!
!
line con 0
line aux 0
line vty 0 4
password password
login
!
!
end
MLPPP Solution Example

The customer has the need to connect but requires increased bandwidth beyond a single E1 or T1 link.
T1 3/1, T1 3/2
192.168.2.x/30
MLPPP
T1 1/1, T1 1/2
192.168.2.x/30
MLPPP
Carrier
ProCurve 7203dl
Hostname: Central
ProCurve 7102dl
Hostname: Remote
Figure 15: MLPPP Solution Example
ProCurve Secure Router 7203dl

!
interface eth 0/1
ip address 192.168.3.254 255.255.255.0
hostname "Central"
no shutdown
no enable password
interface eth 0/2
ip subnet-zero
no ip address
ip classless
shutdown
ip routing
event-history on
interface t1 3/1
no logging email
no shutdown
interface t1 3/2
40
no shutdown
!
!
!
interface t1 3/3
shutdown
!
!
interface t1 3/4
ip route 192.168.1.0 255.255.255.0 192.168.2.1
shutdown
!
!
no ip tftp server
interface t1 3/5
shutdown
no ip http server
no ip snmp agent
interface t1 3/6
no ip ftp agent
shutdown
!
!
!
interface t1 3/7
shutdown
!
!
!
!
interface t1 3/8
shutdown
!
!
!
line con 0
interface ppp 1
no login
ip address 192.168.2.2 255.255.255.252
ppp multilink
line telnet 0 4
no shutdown
login
bind 1 t1 3/1 1 ppp 1
bind 2 t1 3/2 2 ppp 1
End

!
interface eth 0/1
ip address 192.168.1.254 255.255.255.0
hostname "Remote"
no shutdown
no enable password
interface eth 0/2
ip subnet-zero
no ip address
ip classless
shutdown
ip routing
event-history on
interface t1 1/1
no logging email
clock source internal
no shutdown
interface t1 1/2
41
no shutdown
no ip tftp server
no ip http server
interface ppp 1
no ip snmp agent
ip address 192.168.2.1 255.255.255.252
no ip ftp agent
ppp multilink
no shutdown
bind 1 t1 1/1 1 ppp 1
bind 2 t1 1/2 2 ppp 1
line con 0
no login
line telnet 0 4
login
ip route 0.0.0.0 0.0.0.0 192.168.2.2
end
Frame Relay Solution Examples

The customer has the need to connect to multiple sites. Frame Relay provides connectivity across carrier
boundaries and since customer traffic patterns are bursty yet normally low, they chose Frame Relay for
cost reasons.
Frame Relay Solution Example 1

The customer has the need to connect to only a few sites but at great distance crossing carrier networks
to do so. Frame Relay is a good choice for a true private network.
ProCurve 7203dl
Frame Relay
t1 1/1
10.10.1.254/24
Frame Relay
DLCI 254
Ethernet 0/1
192.168.3.253/24
ProCurve 7102dl
t1 1/1
10.10.1.253/24
Frame Relay
DLCI 253
Ethernet 0/1
192.168.1.253/24
Figure 16: Frame Relay Solution Example 1
42
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
interface eth 0/1
ip address 192.168.3.253
255.255.255.0
no shutdown
!
interface eth 0/2
no ip address
shutdown
!
!
interface t1 1/1
no shutdown
!
interface t1 1/2
shutdown
!
!
interface fr 1 point-to-point
frame-relay lmi-type ansi
no shutdown
bind 1 t1 1/1 1 frame-relay 1
!
interface fr 1.254 point-to-point
frame-relay interface-dlci 254
255.255.255.0
!
!
router rip
network 10.10.1.254 255.255.255.0
!
!
no ip tftp server
ip http server
ip snmp agent
no ip ftp agent
!
!
!
line con 0
!
line telnet 0 4
password pnb
!
end
!
!
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
43

!
!
!
interface eth 0/1
ip address 192.168.1.253
255.255.255.0
no shutdown
!
interface eth 0/2
no ip address
shutdown
!
!
interface t1 1/1
no shutdown
!
interface t1 1/2
line-length 0
shutdown
!
!
no shutdown
!
255.255.255.0
!
!
router rip
redistribute static
network 10.10.1.0 255.255.255.0
network 10.10.1.253 255.255.255.0
!
!
!
no ip tftp server
ip http server
ip snmp agent
no ip ftp agent
!
line con 0
login
password pnb
!
line telnet 0 4
login
password pnb
!
end

The customer has the need to connect to only a few sites but at great distance crossing carrier networks
to do so. Frame Relay is a good choice for a true private network. Multiple T1 links are shown to the
central site. The customer may chose to do this for maximizing and equally distributing the bandwidth
available for each site connection. Later they may chose to implement Multi Link Frame Relay.
44
ProCurve 7203dl
t1 1/2
10.10.2.252/24
Frame Relay
DLCI 252
Frame Relay
Cisco 2621
Hostname: 2621_left
ser 0/0
10.10.2.251/24
Frame Relay
DLCI 251
t1 1/1
10.10.1.254/24
Frame Relay
DLCI 254
t1 1/1
10.10.1.253/24
Frame Relay
DLCI 253
Ethernet 0/1
192.168.3.253/24
ProCurve 7102dl
Ethernet 0/1
192.168.1.253/24
FastEthernet 0/0
192.168.2.253/24
!
!
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
interface eth 0/1
ip address 192.168.3.253
255.255.255.0
no shutdown
!
interface eth 0/2
no ip address
shutdown
!
!
no shutdown
!
255.255.255.0
!
no shutdown
!
45
ip address 10.10.2.252 255.255.255.0

no llpd send system-description
llpd send management-address
!
router rip
network 10.10.1.254 255.255.255.0
network 10.10.2.252 255.255.255.0
!
!
no ip tftp server
ip http server
ip snmp agent
no ip ftp agent
!
!
!
line con 0
!
line telnet 0 4
password pnb
!
end
!
!
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
interface eth 0/1
ip address 192.168.1.253
255.255.255.0
no shutdown
!
interface eth 0/2
no ip address
shutdown
!
!
!
interface t1 1/1
no shutdown
!
interface t1 1/2
line-length 0
shutdown
!
!
no shutdown
!
255.255.255.0
!
!
router rip
redistribute static
network 10.10.1.0 255.255.255.0
network 10.10.1.253 255.255.255.0
!
46
!
no ip tftp server
ip http server
ip snmp agent
no ip ftp agent
!
!
line con 0
login
password pnb
!
line telnet 0 4
login
password pnb
!
end
Cisco 2621xm
!
!
version 12.3
service timestamps debug datetime
msec
!
hostname 2621_left
!
boot-start-marker
boot-end-marker
!
!
no network-clock-participate slot 1
no network-clock-participate wic 0
no aaa new-model
ip subnet-zero
ip cef
!
!
!
!
ip address 192.168.2.253
255.255.255.0
duplex auto
speed auto
!
interface Serial0/0
ip address 10.10.2.251 255.255.255.0

encapsulation frame-relay IETF
!
!
ip address 192.168.1.2 255.255.255.0
shutdown
duplex auto
speed auto
!
router rip
network 10.0.0.0
network 192.168.2.0
!
ip classless
ip http server
!
!
control-plane
!
!
line con 0
line aux 0
line vty 0 4
password password
login
!
!
!
end
47

This example shows that true statistical multiplexing can be done across the central site link.
The central site has only one physical link. It may have been determined that both sites only
have about 50% of a T1 requirement or that each of them required access into the central site
at different hours.
ProCurve 7203dl
Frame Relay
Cisco 2621
Hostname: 2621_left
ser 0/0
10.10.2.251/24
Frame Relay
DLCI 251
t1 1/1
10.10.1.254/24
DLCI 254
10.10.2.241/24
DLCI 241
t1 1/1
10.10.1.253/24
Frame Relay
DLCI 253
Ethernet 0/1
192.168.3.253/24
ProCurve 7102dl
Ethernet 0/1
192.168.1.253/24
FastEthernet 0/0
192.168.2.253/24
!
!
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
interface eth 0/1
ip address 192.168.3.253
255.255.255.0
no shutdown
!
interface eth 0/2

no ip address
shutdown
no lldp send systemcapabilities
!
!
!
interface t1 1/1
tdm-group 1 timeslots 1-24
speed 64
no shutdown
!
interface t1 1/2
speed 64
no shutdown
!
!
!
48
no shutdown
!
255.255.255.0
!
255.255.255.0
!
!
router rip
network 10.10.1.254
255.255.255.0
network 10.10.2.241
255.255.255.0
!
!
no ip tftp server
ip http server
ip snmp agent
no ip ftp agent
!
!
line con 0
!
line telnet 0 4
password pnb
!
end
!
!
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
!
interface eth 0/1
ip address 192.168.1.253
255.255.255.0
no shutdown
!
interface eth 0/2
no ip address
shutdown

!
!
!
interface t1 1/1
speed 64
no shutdown
!
interface t1 1/2
line-length 0
shutdown
!
!
!
no shutdown
!
255.255.255.0
!
49
!
router rip
redistribute static
network 10.10.1.0 255.255.255.0
network 10.10.1.253
255.255.255.0
!
no ip tftp server
ip http server
ip snmp agent
no ip ftp agent
!
!
line con 0
login
password pnb
!
line telnet 0 4
login
password pnb
!
end
Cisco 2621xm
!
version 12.3
msec
service timestamps log datetime
msec
!
hostname 2621_left
!
boot-start-marker
boot-end-marker
!
!
no network-clock-participate slot
1
no network-clock-participate wic
0
no aaa new-model
ip subnet-zero
ip cef
!
!
!
!
ip address 192.168.2.253
255.255.255.0
duplex auto
speed auto
!
interface Serial0/0
255.255.255.0
!
!
255.255.255.0
shutdown
duplex auto
speed auto
!
router rip
network 10.0.0.0
network 192.168.2.0
!
ip classless
ip http server
!
!
control-plane
!
!
!
line con 0
line aux 0
line vty 0 4
password password
login
!
!
!
end
50
Multilink Frame Relay Solution Example

This example shows interoperability for Multilink Frame Relay with Cisco 1841.
Cisco 1841
ProCurve 7203dl
Frame Relay
Figure 19: Multilink Frame Relay setup between a Cisco 1841 and ProCurve SR 7203dl
Cisco 1841
Cisco1841#sh run
framing esf
Building configuration...
linecode b8zs
channel-group 0 timeslots 1-24 speed 64
Current configuration : 1381 bytes
version 12.3
interface MFR1
service timestamps debug datetime msec
no ip address
frame-relay multilink bid Cisco1841
hostname Cisco1841
frame-relay intf-type dce
boot-start-marker
interface MFR1.16 point-to-point
boot-end-marker
!
ip address 12.1.1.2 255.255.255.0

mmi polling-interval 60
no mmi auto-configure
no ip address
no mmi pvc
shutdown
mmi snmp-timeout 180
duplex auto
no aaa new-model
speed auto
ip subnet-zero
ip cef
no ip address
shutdown
duplex auto
speed auto
frame-relay switching
interface Serial0/0/0:0
no ip address
encapsulation frame-relay MFR1
no arp frame-relay
controller T1 0/0/0
frame-relay multilink lid Link1-to-7203
framing esf
interface Serial0/0/1:0
linecode b8zs
no ip address
channel-group 0 timeslots 1-24 speed 64
encapsulation frame-relay MFR1
no arp frame-relay
controller T1 0/0/1
frame-relay multilink lid Link2-to-7203
51
ip classless
line con 0
ip http server
line aux 0
line vty 0 4
login
control-plane
end

ProCurveSR7102dl#sh run
no shutdown
hostname "ProCurveSR7102dl"
no enable password
frame-relay multilink
ip subnet-zero
no shutdown
ip classless
ip routing
event-history on
no logging email
ip address 12.1.1.1 255.255.255.0

!
!
!
interface eth 0/1
no ip address
no ip tftp server
shutdown
no ip http server
interface eth 0/2
no ip snmp agent
no ip address
no ip ftp agent
shutdown
line con 0
no login
interface t1 1/1
!
line telnet 0 4
no shutdown
!
login
!
interface t1 1/2
end
ADSL Solution Example

Note: Please note that these examples, and those in further sections, are given for your study and consideration only. They
are to help you reach a better understanding of the fundamental concepts before configuring your own application. It will
be necessary for you to modify these examples to match your own network design.
ADSL Solution Example

The customer needs to access the Internet through the ADSL modules.
ADSL Solution Example with PPPoE over ATM over ADSL

This example shows the customer needs basic connectivity to the Internet and the Service
Provider has a packet routing core. PPPoE establishes a connection to that router over the
carriers ADSL link and ATM layer 2.
52
Service Provider
ProCurve SR 7102dl
Figure 20: ADSL Solution Example with PPPoE
!
hostname "ADSL-Router"
!
ip routing
!
!
ip dhcp-server pool "pool-forlan"
network 192.168.1.0
255.255.255.0
domain-name "yourdomain.com"
dns-server x.x.x.x y.y.y.y
netbios-node-type h-node
default-router 192.168.1.1
lease 1
!
interface eth 0/1
255.255.255.0
no shutdown
!
!
interface eth 0/2
no ip address
shutdown
!
!
interface adsl 1/1
training-mode multi-mode
no shutdown
!
!
interface atm 1 point-to-point
no shutdown
bind 1 adsl 1/1 atm 1
!
!
interface atm 1.1 point-to-point
no shutdown
pvc 8/35
no ip address
!
!
!
interface ppp 1
ip address negotiated
no fair-queue
53
ppp chap hostname id-given-byisp

ppp chap password pw-given-byisp
no shutdown
bind 2 atm 1.1 ppp 1 pppoeclient
!
!
!
ip access-list extended lan-acl
permit ip any any

!
!
ip policy-class FROM-LAN
nat source list lan-acl
interface ppp 1 overload
!
!
end
ADSL Solution Example with RFC 1483

The customer needs to bridge on the ATM/ADSL interface into the service provider. RFC 1483
support gives the customer the ability to bridge traffic over the ADSL/ATM interface and still
forward that traffic to its Ethernet interface.
Note: Layer 1 (Physical) is ADSL (interface ADSL 1/1), Layer 2 (Link Layer) is ATM (interface ATM 1) and is bound to ADSL
1. Layer 2.1 is on top of ATM where there is an adaptation layer AAL5MUX or AAL5SNAP that we define for the ATM
interface
Data in ATM is sent in a Permanent Virtual Circuit (PVC) defined with its VPI / VCI and is bound to the Point to point ATM
interface named ATM 1.1
Layer 2.2 is the PPP Layer as defined in logical interface PPP 1 that used
hostname "Secure-Router"
interface eth 0/2
no ip address
ip routing
shutdown
interface adsl 1/1
ip firewall
no shutdown
interface atm 1 point-to-point
no shutdown
bind 1 adsl 1/1 atm 1
ip dhcp-server excluded-address
192.168.1.1
!
interface atm 1.1 point-to-point
no shutdown
ip dhcp-server pool "pool-for-lan"
pvc 8/36
network 192.168.1.0 255.255.255.0

domain-name "yourprovider.com"
dns-server x.x.x.x, y.y.y.y
encapsulation aal5mux ip
ip address dhcp
bandwidth 576
netbios-node-type h-node
default-router 192.168.1.1
lease 1
!
ip access-list standard manage-rtr

permit 192.168.1.0 0.0.0.255
ip access-list extended lan-acl
remark used for Nat
interface eth 0/1

ip address 192.168.1.1 255.255.255.0
access-policy FROM-LAN
no shutdown
permit ip any any

!
ip policy-class FROM-LAN
allow list manage-rtr self
54
nat source list lan-acl interface atm 1.1

overload
!
end
Feature Set Consideration for ProCurve Secure Router 7000dl Series

The ProCurve Secure Router 7000dl series routers are the 7102dl and the 7203dl as
of this writing. Check http://www.procurve.com for further specifications.
The ProCurve Secure Router 7000dl series supports the following physical interfaces
as of this writing:
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
J8451A ProCurve SR dl 1xT1 module

J8452A ProCurve SR dl 1xT1 + DSX-1 module
J8453A ProCurve SR dl 2xT1 module
J8454A ProCurve SR dl 1xE1 module
J8455A ProCurve SR dl 1xE1 + G.703 module
J8456A ProCurve SR dl 2xE1 module
J8458A ProCurve SR dl 1-port Serial module
J8459A ProCurve SR dl 1xADSL2+ /A module
J8759A ProCurve SR dl 1xADSL2+ /B module
J8460A ProCurve SR dl 1-port ISDN BRI U backup module
J8461A ProCurve SR dl 1-port ISDN BRI S/T backup module
J8462A ProCurve SR dl Analog Modem backup
J8463A ProCurve SR dl Wide 8xT1/E1 module
J8755A ProCurve SR Serial Cable V.35 DTE
J8757A ProCurve SR Serial Cable X.21 DTE
The ProCurve Secure Router 7000dl series supports the following Layer 2 WAN
protocols as of this writing:
o
o
o
o
o
o
o
o
HDLC Cisco compatibility

PPP
Multilink PPP
Frame Relay (FRF.12)
Multilink Frame Relay (FRF.16)
ATM
PPPoE
PPPoA
Early releases of the ProCurve Secure Router 7000dl series did not allow the HDLC
link to initiate the backup module upon failure. Options for deployment include:
Using PPP or Frame Relay if there is requirement to monitor the data

link layer.
Apply the backup to the physical interface.
The current implementation of ATM only support AAL 5 or UBR which is acceptable for
IP data traffic.
Check http://www.procurve.com for the latest free software update, or for additions
to the physical interfaces supported.
How to Use the ProCurve Secure Router 7000dl Series

Use the 7102dl at the branch offices that primarily need access to their central site.
If the branch needs to increase bandwidth out from the site, consider using a single
7203dl in a stacked configuration along with one or more of the suggestions below for
increasing bandwidth.
Use a single 7203dl at the central sites of networks not requiring greater throughput
than 12 Mbps or do not require more connectivity than the equivalent of 12 E1 or T1

interfaces.
55
Use multiple 7203dl at central sites requiring greater throughput or connectivity.
Stacking 7203dl routers at the central site is possible. Properly designed this
stacked solution can enhance overall network availability since a single device
failure, instead of a stack, could take down the entire network. Multiple devices are
sometime preferable for high availability.
Backup for the 7102dl can be either through an optional backup module, or through
use of the Internet as a backup route and the use of a floating static route.
See High Availability section of this paper for details.
Increasing bandwidth for the primary routes may be accomplished with one or more
of the following:
Use of serial interface over a single T1 or E1 that connects into a carrier

supplied CSU/DSU that has greater bandwidth than a singe T1 or E1 can
supply. Often these special devices are region and location specific
and may not be available in all areas. Ask your carrier or service
provider.
o
o
Multilink PPP or Multilink Frame Relay

Limit your Multilink implementation to a maximum of 8 physical links.
An Overview of IP, Static and Dynamic Routing Protocols

Executive Summary
For the interconnection of IP networks over a WAN the routers must have these basic elements:
A routable protocol, in this case IP.

A means to inform the routers of the proper path to forward the IP packet along, such
as with static routes or dynamically with RIP or OSPF.
The ability to bridge other routable protocols, such as DecNet or IPX.

Overview
Routers perform many functions; mainly they connect multiple networks together. Routers can
function as LAN or WAN devices, with the ability to interconnect both local area and wide area
networks. When multiple logical networks must communicate or be interconnected, the
network requires a router. One of the routers most prominent functions is to forward Layer 3
packets based upon logical addresses. For the WAN router, another critical feature that is often
overlooked is the complete replacement of Layer 2 header information and protocol when
forwarding a packet from one network to another.
A router examines and learns the source layer 3 address of each packet. Its routing table is
contained in memory and used for these learned addresses. Routers determine the best path of
each packet through consulting the routing table and using routing metrics. The router can
then forward packets from one network to another.
Routing metrics are based upon criteria such as speed of the link, number of points between
final end points, or other such factors. Customers may choose to administrate this routing
decision manually, as with static routes, or dynamically, as with dynamic routing protocols
such as RIP and OSPF. Typically smaller installations, those less than approximately 5 sites, will
choose to control the paths that forwarded IP packets take through the use of static routes.
Another common usage for static routes is when there is a star topology where all the remote
sites go directly back to the central site. These simple topologies or smaller installations can
use static routing effectively. On the other hand static routing can be administratively
burdensome with more than a handful of sites so dynamic routing protocols are chosen for lager
installations. All routers should interoperate with each other using a standard routing protocol
such as RIP or OSPF. There are other proprietary routing protocols such as IGRP or EIGRP, but
installations today that need interoperability between protocols choose OSPF for speed and
efficiency. If a router has a proprietary routing protocol it will also support a number of
standards based protocols also. Customers can, and do, effectively use these together in their
networks.
IP General
An IP network is a collection of IP hosts that share a common network number. The network
portion of each hosts IP address is defined by a mask, which is often called the subnet mask.
The concept of a subnetwork, subnet, or subnet mask is related to classful IP addressing.
The 32-bit IP address space was originally carved up into several classes. Each class has a
natural mask that initially determines what portion of the address represents the network and
56
what portion represents the host. The picture below shows the class divisions and the resulting
natural masks.
Class A network numbers are between 0.0.0.0 and 127.0.0.0 (network 0

is undefined and network 127 is reserved) with a natural mask of 8 bits.
0.0.0.0
127.255.255.255
00000000 00000000 00000000 00000000

01111111 11111111 11111111 11111111
Class B network numbers are between 128.0.0.0 and 191.255.0.0,

with a natural mask of 16 bits.
128.0.0.0
191.255.255.255
10000000 00000000 00000000 00000000

10111111 11111111 11111111 11111111
Class C network numbers are between 192.0.0.0 and 223.255.255.0,

with a natural mask of 24 bits.
192.0.0.0
223.255.255.255
11000000 00000000 00000000 00000000

11011111 11111111 11111111 11111111
Table 7: IP address classes and natural masks
The three IP address classes can be summarized as follows:
When the first octet of an IP address is between 1 and 126, the address falls within
the Class A range, with a natural mask of 8 bits. The 8-bit mask results in a range of
over 16 million addresses for each of the 126 Class A network numbers.
the Class B range, with a natural mask of 16 bits. The 16-bit mask results in a range
of 65,536 addresses for each of the 16,384 Class B network numbers.
the Class C range, with a natural mask of 24 bits. The 24-bit mask results in a range
of 256 addresses for each of the 2,097,152 Class C network numbers.
IP networks and subnetworks

The rules of classful IP addressing dictate that the mask associated with each address can be
made longer than the natural mask.
Classful addressing does not allow you to use a mask that is shorter than the natural mask. If
you are using public addresses within an enterprise, this is a reasonable restriction, because the
use of a shorter mask will overlap with the addresses assigned to other enterprises. However,
particularly when using private addressing, there can be a valid use for masks that are shorter
than the natural mask (sometimes called a "supernet mask").
If you are using RIP version 1 as route exchange protocol, you need to comply with the classful
addressing restriction mentioned above.
Classful vs. classless IP addressing

When you are using an IP route exchange protocol that supports classless IP addressing, such
as RIP version 2 or OSPF, every range is defined by a starting address and a mask. The mask
may be expressed in dotted-decimal format or as a count of the number of consecutive "1" bits
(often referred to as "CIDR" notation). Although you may see or hear references to a "subnet"
mask, the term "subnet" is actually relevant only in a classful environment. Consider the
following comparison:
In classful addressing, the two address ranges 10.1.1.0/24 and 10.1.2.0/24 have an inherent
relationship as portions (sometimes called "children") of the same classful (sometimes called
"parent") network 10.0.0.0.
57
With classless addressing, the mask defines the boundaries of each address range; the value in
the first octet has no effect on determining the size of the address range.
IP routing
The primary role of an IP router is to interconnect networks and facilitate the forwarding of IP
traffic between hosts. The destination of any particular packet may be a unicast address in
the range 1.0.0.0 through 223.255.255.255, or a multicast address in the range 224.0.0.0
through 239.255.255.255.
A router uses a variety of methods to gather information about networks, which are groups or
ranges of IP addresses. The router stores this information in a table and uses the entries to
forward traffic toward its ultimate destination.
Remote networks
When some of the networks are reached through a different router, there must be a way to
share information between routers and facilitate the movement of customer traffic between
remote networks.
Each router records in its route table the existence of its directly connected networks. However,
in order to dynamically acquire information about non-local networks each router needs to
receive information from neighboring routers.
Route exchange protocols define rules for exchanging information. These rules may specify:
The format of the information to be exchanged,

conditions that trigger the transmission of information,
procedures for adding new information to the routing table, and
procedures for replacing or removing inaccurate information.
Routing Information Protocol (RIP)

The simplest route exchange protocol is the Routing Information Protocol. When RIP is applied
to an IP interface, the router sends periodic updates (every 30 seconds) to neighboring routers.
RIP is a distance vector protocol, which means that its updates include the distance, measured
in the number of hops, between a particular router and a network.
Static routes
Another way to administratively intervene on path selection is to create a static route that
specifies a next hop that leads to the more desirable path. While cost is used for path selection
when all choices come from the same source, such as RIP, cost alone is not used as tie-breaker
when routing information comes from different sources. By default, statically defined routes are
preferable over RIP-learned routes.
Administrative distance
Static routes rely upon information supplied by an administrator; they do not use RIP to
dynamically gather information. Static routes thus constitute a different source of information.
Most routers use administrative distance to make path selection choices when routing
information comes from different sources. You can think of administrative distance as a
"reliability" or "believability" factor, where the lower number is considered more reliable or
closer.
All route exchange protocols and sources of routing information, including static routes, have a
default administrative distance. RIP-learned routes have a default administrative distance of
120, which may be modified. Static routes have a default administrative distance of 1, which
makes them preferable over RIP routes. The administrative distance for a static route may be
specified when the route is defined.
When static routes are intended to replace dynamically learned routes, the administrative
distance may be left at the default value of 1. However, another choice for administrative
intervention would be to use a static route as a backup to a RIP-learned route.
Default static route

A commonly defined static route is the default static route 0.0.0.0 with a mask of 0.0.0.0. The
default route is used to minimize the size of routing tables, thus maximizing router
performance.
For example, in networks that are connected to the public Internet, it would be impractical for
the routers to maintain routing table entries for every network available worldwide. Instead, the
58
routers typically keep detailed information only about networks within their own domain or
"autonomous system".
By using the default route, traffic destined for networks not listed in the routing table is
forwarded toward a router with more specific information. Typically, there are several levels of
default routes that ultimately lead to a "default-free zone," where detailed Internet routing
information is available.
Default routes are also used in private enterprise networks to minimize route table size. Some
remote areas may maintain detailed information only about their site, and use the default route
to represent any destination not specifically listed in the route table.
The existence of the default route does not change the lookup process. However, since the
default route specifies an address range that matches with EVERY destination address, some
packets will have multiple matches. In this case, the MOST SPECIFIC MATCH is chosen. This is
the match with the longest mask.
The primary value of the default route is that, since it matches with every packet, it provides a
"next hop" for packets that are destined for unknown networks.
Open Shortest Path First (OSPF)

OSPF is a link state route exchange protocol that was standardized by the Internet Engineering
Task Force to support scaleable, resilient networks. A primary goal of OSPF is to reduce the
frequency of update traffic. Another goal is fast convergence. Meeting these two goals results
in a tradeoff, that OSPF consumes more memory and CPU resources on the router than distance
vector protocols.
RIP has a diameter limit due to the way that information is passed between routers. RIP routers
pass on to neighbors their own view of the network every 30 seconds. The neighbor, in turn,
interprets the information from its own perspective and passes it on again. One inevitable
result is slow convergence. Another possible outcome is that this "rumor-based" information can
be misinterpreted and disrupt connectivity.
Every OSPF router generates an advertisement that includes the identities of its active IP OSPF
interfaces, including address, mask, and cost. The advertisements are immediately flooded,
without being changed, by other OSPF routers and this results in very fast convergence. All
OSPF routers within a defined domain (known as an area) receive exactly the same
advertisements.
OSPF routers collect the advertisements in a database and each independently calculates the
lowest cost (shortest) path to each network within the domain. It does this by examining each
element in the database and building a tree, with itself as the root. The advertised networks
are either branches or leaves of the tree, joined by the advertising routers.
From the shortest path tree, each router derives the "next hop" for each destination network,
and places that information in its routing table. The maximum path cost is over 65,000.
If any OSPF router experiences a state change, i.e. an OSPF interface goes up or down, it
generates a new advertisement. The advertisement is flooded throughout the domain. Each
router replaces the obsolete advertisement with the latest information, and rebuilds the tree,
based on the very latest information. The resulting routes replace the ones in the routing table.
It is important to recognize that, although advertisements are flooded by OSPF routers, this
does not mean that IP customer traffic is flooded. The IP forwarding process used by OSPF
routers is the same as that used by RIP routers. IP traffic is always forwarded one "hop" at a
time.
OSPF hierarchy
An area is a collection of contiguous networks, identified by its area ID. The area ID may be
expressed as a decimal number or a dotted-decimal number, depending on router vendor and
customer preferences. The ProCurve Secure Router 7000dl Series supports both formats.
The definition of multiple areas introduces an additional level of hierarchy that makes it possible
to simplify routing tables by hiding the topology of some portions of the network. The highest
level of hierarchy is an Autonomous System (AS), which may be defined as a group of
networks under common administrative control. An OSPF AS consists of one or more areas.
Every OSPF router is uniquely identified by a router ID.
Each OSPF area is made up of two basic types of networks:
59
Networks that have only one router are recognized as stub networks. A stub network
is a destination address range, but is like a leaf on the tree because it does not serve
as a path to any other network.
Networks that have multiple connected routers are recognized as transit networks.
Transit networks provide potential paths to other destinations, like the branches of a
tree.
The flow of link state information

While routers using RIP send updates that contain the entire content of their route table to
neighbors, OSPF routers send information about their directly connected links.
A link state advertisement (LSA) is a unit of information that is generated by an OSPF router
and flooded through the OSPF domain. There are six types of LSAs; some are flooded
throughout an area, while others cross the boundaries between areas and are flooded
throughout the entire AS.
Each router collects LSAs in a link state database. Due to the flooding of LSAs, all routers in the
same domain have an identical link state database. Routers in different areas will have different
link state database entries.
Each router runs a link state algorithm, called the Dijkstra algorithm, using the information in
the link state database as input. The output of the link state algorithm is a shortest path first
tree that maps out a loop-free path to every known network and router.
Multiple OSPF areas

OSPF was designed to scale to very large networks. To keep the database at a manageable size,
and to minimize the area that is affected by interface state changes, OSPF networks are often
divided into multiple areas.
Because router LSAs and network LSAs flow freely within an area, each router in that area has
detailed, consistent information in its link state database. Routers that have all of their
interfaces in the same area are known as internal routers their link state database entries all
relate to the same area. Routers that connect to more than one area are known as area border
routers (ABR).
An area border router maintains a link state database for each the areas it interconnects. The
area border router transfers information between areas by generating a third type of LSA, the
summary LSA, which reports the networks in one area into another area. Network and router
information from one database is converted into summary LSAs and copied into the other
database.
In addition to the reduction in the size of the link state database for internal OSPF routers,
dividing the internetwork into multiple areas reduces the impact of link state changes. Since
the scope of router LSAs and network LSAs is a single area, those advertisements are not
flooded between areas. For example, if a link state change occurs in area 0.0.0.1, the routers
in other areas do not need to run the link state algorithm and recalculate routes. This is
because routers internal to an area represent networks from other areas as leaf objects in the
SPF tree.
To minimize the number of advertisements that are sent between areas, administrators often
configure network summaries. One can summarize four networks with 24-bit masks into a
single advertisement with a 22-bit mask.
Summarization can help to minimize the number of summary LSAs that are sent between areas
and, as a result, minimizes the number of entries in the routing table.
Autonomous System Boundary Router (ASBR)

While an OSPF autonomous system (AS) is often made up of multiple areas, information
regarding non-OSPF networks is considered "external" to the OSPF AS. OSPF routers that have
learned route information from sources other than OSPF are known as autonomous system
boundary routers (ASBRs). RIP routes, static routes, and directly connected networks nonOSPF interface, are some examples of external information.
ASBRs are configured to redistribute non-OSPF routes into the OSPF domain using type 5 LSAs,
known as external LSAs. To cause an ASBR to generate external LSAs to represent RIP and
static routes, enter the following command within the OSPF configuration level:
60
The ASBR generates one external LSA for each non-OSPF network, floods the LSAs to its
neighbors, who flood the LSAs, unchanged, to their neighbors. As a result, all of the routers in
the area know about the external networks.
To minimize the number of LSAs, the external address space may be summarized using the
summary-address command within the OSPF configuration level. More on route
summarization at the end of this section.
External LSAs flow over area boundaries. When an area border router receives type 5 LSAs
over an adjacency with a router in a non-backbone area, it copies them into the backbone area
link state database and floods them over adjacencies with routers in the backbone. External
LSAs that it receives over adjacencies with other backbone routers will be flooded into nonbackbone areas only if they are normal areas.
Normal areas and stub areas

While type 5 LSAs flow freely between normal areas, they are not copied into "stub" areas.
Instead, the area border routers inject the default route to represent external (non-OSPF)
networks. The practical effect of defining a stub area is to reduce the number of entries in the
link state database of the internal routers in the stub area, resulting in lower memory
consumption and a shorter time to run the link state algorithm.
The route table within a stub area will have an entry for each network in the area, a summary
of the address space in each of the other areas, and a default route that represents external
(non-OSPF) networks. The backbone cannot be configured as a stub area.
Not-so-stubby areas (NSSA)

Stub areas are an important tool in minimizing the size of the link state database and the route
table within non-backbone areas. However, a stub area cannot contain any ASBRs; type 5 LSAs
cannot exist within the link state database of routers in a stub area.
The not-so-stubby area (NSSA) allows ASBRs to exist within an area that uses the default route
to reach non-OSPF networks in other areas.
The NSSA combines the benefits of the stub area with some of the flexibility of a normal area.
The ASBR within the NSSA advertises the external networks as type 7 LSAs, which is a legal
LSA type within the NSSA. The ABR connected to the NSSA converts the type 7 LSAs to type 5
LSAs and floods them into the backbone area, where they are handled like any other type 5
LSA. ABRs that connect stub areas to the backbone will block the external information and
inject the default route to spare the internal routers from tracking the details about non-OSPF
destinations.
Border Gateway Protocol

When one looks at the Internet from the perspective of a customer, it appears as a collection of
resources that you can access through an ISP. The underlying topology of the Internet, as well
as the processes that enable communication between distinct entities all over the world, are
irrelevant from the customer perspective. However, in order to understand the operation of
Border Gateway Protocol (BGP) it is useful to understand a few things about Internet topology
and communication between different enterprises.
61
Transit
AS
Transit
AS
Non-Transit
AS
Transit
AS
Non-Transit
AS
Non-Transit AS
(multi-homed)
Non-Transit
AS
Figure 21: Types of Autonomous Systems in the Internet
In its current form, the Internet is an arbitrary collection of autonomous systems. Exterior
gateway protocols are used for communication between autonomous systems, and their use
obscures the interior topology of an AS in such a way as to make it irrelevant what route
exchange protocols are used within.
An enterprise that uses Border Gateway Protocol (BGP) to exchange internal address space
information with an ISP must have its own AS number. The connection between a router in
one autonomous system and another router in a different autonomous system is called an
external BGP connection.
BGP operates over the connection-oriented transport layer protocol TCP. The BGP routers are
each configured with a peer session that specifies the IP address and AS number of the other,
and they form a connection that remains up indefinitely.
Routers that are configured to support BGP are called BGP speakers. Two BGP speakers
communicate over a peer session. The BGP router on the other side of a peer session is called
a BGP neighbor.
External BGP operation

Once the connection has been established over the external BGP peer session, BGP routers send
incremental updates that include summarized address ranges and AS numbers. They also send
keepalives to maintain the session. All BGP messages are sent to TCP port 179.
Network Layer Reachability Information
A route is not a network or subnet as in the previous route exchange protocols, but is a unit of
information that pairs a destination with path attributes.
62
150.10.0.0/16
AS_PATH: 400 200
AS 100
198.155.70.0/24
AS_PATH: 300 200
AS
200
198.155.70.0/24
AS_PATH: 300
AS 300
198.155.70.0/24
150.10.0.0/16
AS_PATH: 400
AS 400
150.10.0.0/16
Figure 22: Network Layer Reachability Information
A destination is a range of IP addresses that are reported using prefix/length notation, where
the prefix is the starting address of the range and length defines the size or boundaries of
the range. For example, the destination 150.10.0.0/16 is a range consisting of the 65,536 (216)
IP addresses between 150.10.0.0 and 150.10.255.255.
Path attributes, known in the BGP specification as AS_PATH, is a list of the autonomous systems
through which a route passes. BGP4 routers can use this list of traversed autonomous systems
to detect and eliminate routing loops.
Internal BGP
The ProCurve Secure Router 7000dl series supports both Internal iBGP and External eBGP.
While there are many similarities between external and internal BGP, the most important
difference is that the BGP speakers in an Internal BGP peer session are in the same AS.
Internal BGP is used within a transit AS, as is shown in the diagram below. Please check
http://www.procurve.com for free software updates due soon that will include further
enhancements to BGP.
63
eBGP
eBGP
AS 500
AS 100
AS 200
iBGP
iBGP
Figure 23: Internal BGP
If an AS has multiple BGP speakers, it could be used as a transit service for other ASs. As you
can see in the iBGP diagram, AS 500 is a transit AS for AS 100 and AS 200.
It is necessary to ensure reachability within an AS before sending the information to an external
AS. This is done by a combination of internal BGP peering between router inside an AS and by
redistributing BGP information to Internal Gateway Protocols (IGPs) running in the AS.
When BGP is running between routers belonging to the same AS, it is call iBGP. When BGP is
running between routers that are in different ASs, it is called eBGP.
T1 3/1, T1 3/2
192.168.2.x/30
MLPPP
T1 1/1, T1 1/2
192.168.2.x/30
MLPPP
ISP
ProCurve 7203dl
Hostname: Central
ProCurve 7102dl
Hostname: Remote
AS 65300
Figure 24: BGP Example

!
hostname "Central"
ip subnet-zero
no enable password
ip classless
64
ip routing
shutdown
event-history on
interface ppp 1
ip address 192.168.2.2 255.255.255.252
no logging email
ppp multilink
no shutdown
bind 1 t1 3/1 1 ppp 1
bind 2 t1 3/2 2 ppp 1
interface eth 0/1
ip address 192.168.3.254 255.255.255.0
router BGP 65300
no shutdown
no auto-summary
no synchronization
interface eth 0/2
network 192.168.2.0 mask

255.255.255.252
no ip address
shutdown
neighbor 192.168.2.1
default-originate
soft-reconfiguration inbound
!
!
remote-as 65300
interface t1 3/1

no shutdown
!
!
!
!
interface t1 3/2
no shutdown
no ip tftp server
no ip http server
no ip snmp agent
!
interface t1 3/3
shutdown
no ip ftp agent
!
!
!
interface t1 3/4
shutdown
!
!
!
!
interface t1 3/5
shutdown
!
interface t1 3/6
shutdown
!
interface t1 3/7
shutdown
!
!
!
line con 0
no login
!
line telnet 0 4
login
!
End
interface t1 3/8

Remote#sh run
no enable password
!
ip subnet-zero
ip classless
hostname "Remote"
ip routing
!
65
event-history on
bind 1 t1 1/1 1 ppp 1
bind 2 t1 1/2 2 ppp 1
no logging email
router BGP 65300
no auto-summary
no synchronization
!
!

255.255.255.252
interface eth 0/1
neighbor 192.168.2.2
no default-originate
ip address 192.168.1.254 255.255.255.0
no shutdown
remote-as 65300
!
interface eth 0/2
no ip address
shutdown
!
!
!
!
no ip tftp server
!
!
no ip http server
interface t1 1/1
no ip snmp agent
no ip ftp agent
no shutdown
interface t1 1/2
no shutdown
interface adsl 2/1
line con 0
shutdown
no login
!
line telnet 0 4
!
interface ppp 1
ip address 192.168.2.1 255.255.255.252
ppp multilink
login
!
end
no shutdown
Classless InterDomain Routing (CIDR)
It is within the BGP environment that the benefits of Classless InterDomain Routing (CIDR)
become apparent. CIDR was suggested in the late 1980s and then later mandated by the
Internet Engineering Task Force in the early 1990s when the number of networks attached to
the Internet started to increase rapidly.
The diagram below illustrates the addressing relationship between an ISP, its subscribers, and
other ISPs. Recall that in classless addressing, a starting address followed by a mask defines a
range address whose size is some power of 2, regardless of the value in the first octet of the
address.
In the example, ISP A owns the range of addresses 202.0.0.0/14, provides Internet service to
1,024 businesses. This means that 262,144 (218) IP addresses between 202.0.0.0 and
202.3.255.255 are all reachable through ISP A. This ISP allocates a range of 256 addresses to
each of its subscribers. Each of these address ranges is equivalent to a Class C network.
66
When ISP A advertises its address range to another service provider, ISP C, Classless
InterDomain Routing allows the service provider to summarize the address range using a mask
that breaks the rules of classful addressing because it is shorter than the natural mask.
If classless addressing were not allowed, this service provider would have to individually
advertise 1,024 Class C network numbers (each with a classful 24-bit mask), beginning with
202.0.0.0, 202.0.1.0, 202.0.2.0, and continuing on through 202.3.255.0.
Given the fact that the total 32-bit IP address space allows over 2 million individual IP
addresses, the ability to aggregate address space beyond classful boundaries minimizes the
number of IP address ranges in the core of the Internet.
In the diagram below, ISP B advertises another address range - 202.4.0.0/14.
advertise:
202.0.0.0/14
ISP A
subscribers
advertise:
202.4.0.0/14
202.0.0.0/24
202.0.1.0/24
202.0.2.0/24
...
202.3.255.0/24
ISP B
subscribers
address ranges that do

not follow classful
boundaries
202.4.0.0/23
202.4.2.0/26
202.4.2.64/26
202.4.2.128/26
...
202.7.255.0/24
ISP C
ISP B has over 1,000 business

Internet customers and owns a
pool of IP addresses between
202.4.0.0 and 202.7.255.255.
Some of its customers need as
many as 512 addresses, others
need as few as 64 addresses.
CIDR allows ISP B to advertises
its entire range of addresses
(262,144 or 218) rather than 1,024
Class C addresses.
Figure 25: Classless InterDomain Routing ISP B
As was true with ISP A, a range of 262,144 (218) IP addresses between 202.4.0.0 and
202.7.255.255 are reachable through ISP B. If constrained to classful addressing, this service
provider would also have to individually advertise 1,024 Class C network numbers, beginning
with 202.4.0.0, 202.4.1.0, 202.4.2.0, and continuing on through 202.7.255.0.
Notice that ISP Bs business customers have different size networks and thus require different
address range sizes. In the diagram, some of the subscribers have an address range that uses
a 26-bit mask and other subscribers use a 23-bit mask. Customers that have address ranges
with a mask shorter than the 24-bit mask allowed by classful IP addressing may need to use
classless internal route exchange protocols as well. RIP version 1 follows classful addressing
rules, which means that an IP address in the range between 192.0.0.0 and 223.255.255.255
cannot have a mask shorter than 24 bits.
Classless addressing defines address ranges by specifying a starting address and a mask whose
length is not restricted by the value in the first octet. An address range that has a 23-bit mask
contains 512 addresses, and any address range that has a 26-bit mask contains 64 addresses.
When ISP C advertises this range of addresses to other providers, it can aggregate the address
space into a single statement, which is shown below. The address range 202.0.0.0/13, defines
a range of 524,288 addresses (219) between 202.0.0.0 and 202.7.255.255. If we were
referring to this block of addresses in classful terms, there would be 2,048 Class C networks
(211) between 202.0.0.0 and 202.7.255.0. CIDR minimizes the number of advertisements that
are required to advertise this address space.
67
advertise: 202.0.0.0/13
ISP C
advertise:
202.0.0.0/14
advertise:
202.4.0.0/14
ISP B
ISP A
subscribers
202.0.0.0/24
202.0.1.0/24
202.0.2.0/24
...
202.3.255.0/24
ISP A and ISP B forward their traffic

toward ISP C, who in turn forwards it
on toward the Internet core.
When ISP C reports the address
spaces owned by ISP A (262,144
addresses or 218) and ISP B
(262,144 or 218), it aggregates the
address ranges in a single statement
as 202.0.0.0/13 (524,288 or 219).
subscribers
202.4.0.0/23
202.4.2.0/26
202.4.2.64/26
202.4.2.128/26
...
202.7.255.0/24
Figure 26: Classless InterDomain Routing ISP C
The true benefit of CIDR is found in the "default-free zone" of the Internet. This is the top level
of the hierarchy, a place that is sometimes called the "core". In every other part of the
Internet, as well as in private enterprises, routers often make use of a "default route" that is
used for forwarding traffic toward destinations whose addresses are not specifically known.
However, every address range is represented in the default-free zone of the Internet; there are
no default routes.
The developers of the CIDR concept saw it as a way to limit the complexity of route selection
within the default free zone by limiting number of "routes" or address ranges. Without the
classless address space aggregation that CIDR allows, there would be potentially over 2 million
Class C addresses, 16,384 Class B addresses, and 126 Class A addresses, rather than the
approximately 60,000 routes there are today.
Bridging and Routing at the same time. Caveats. The issue is when you have multiple layer 3 protocols and IP on the
same remote network. Do I discuss or mention we will have it in future releases.
Overlapping address spaces and the longest mask rule

Though it is possible to encounter overlapping address spaces when using a single IP route
exchange protocol, the use of multiple route exchange protocols increases the likelihood.
When the lookup process encounters multiple matches, it places the next hop for the MOST
SPECIFIC match into the address cache. This is easiest to understand if you think of routing
table entries as address ranges, rather than networks or subnets.
The most specific match rule is also known as the longest mask rule. If there is any match
other than the default route (starting address 0.0.0.0 with a 0-bit mask), it will be the one used
to route the traffic in question.
We will consider a few path selection examples. The following address spaces exist in the
routing table:
10.10.15.0/24
10.10.0.0/16
10.0.0.0/8
0.0.0.0/0
In path selection example 1 a packet comes through with the destination address 10.10.15.12.
This packet matches with all 4 address ranges, but the most specific match for that destination
address is 10.10.15.0/24, which is a range of 256 addresses.
68
This packet matches with the last two address ranges. The most specific match is 10.0.0.0/8,
which is a range of over 16 million addresses.
The most specific (and only) match is the default route (0.0.0.0).

IP version 4 overview of IP networks
Dynamic routing protocols include,
There are both classful and classless
RIP is an early distance vector
and subnetworks as it is still the

primary addressing scheme used for
by most companies world wide.
IP addressing schemes.
Route tables are used by routers to

forward IP data from network to
network
Routers make decisions about
forwarding IP through configuration

of static routes or dynamic routing
protocols.
Static routes can be used to
manually administer routing
Routing Information Protocol (RIP),

a distance vector protocol, and
OSPF, a link state protocol.
dynamic routing protocol, version 2
allows for CIDR
OSPF version 2 is a link state

dynamic routing protocol.
Route summarization is a technique
to help control the size of routing

tables in routers and also control the
amount of routing protocol traffic
across WAN links.
BGP is an external gateway protocol

for interconnecting private
autonomous systems.
How These Technologies are Used

Use static routes for simpler
topologies and smaller networks.
Dynamic for multiple routing choices
All of these protocols are used for
the maintaining routing information

for forwarding of IP data.
to make from point A to point B.
Some dynamic routing protocols
require multicasts, OSPF and RIP

version 2. In order for these
protocols to be used across the
Internet, while IPSec VPNs are used,
GRE will also need to be
implemented. (IPSec VPN and GRE
are discussed later in this paper).
One of the most common techniques for controlling routing updates is the use of route
summarization. It is used to both minimize the processing of routing data and to control traffic
on a WAN. This paper summarizes the technique in a separate appendix at the end of this
paper.
Advantages
The primary advantage of the dynamic over static routing protocols is that they update routing
information based upon current route states.
The primary advantage of OSPF over RIP is that OSPF is more responsive than RIP. OSPF also
allows for a greater diameter network and have greater flexibility for integration.
OSPF can be used effectively in both small and large networks. One key to remember is that
for small networks you do not need to have multiple areas. Everything will work quite well on
the backbone area.
69
Disadvantages
RIP is slower to converge. This means that if the network has become temporarily unstable,
due to link outages or other reasons, network users will gain access to resources as quickly as
with OSPF or Static routes.
OSPF is CPU and memory intensive in very large networks but this can be adjusted for with
proper use of route summarization. Much of this also depends on the complexity and scale of
the network.
What to Determine During Planning or for Implementation
Addressing Schemes
Ways to take advantage of summarization
Solution Examples for Layer 3

OSPF Solution Example 1
The customer has many customers that access their core network at Area 0. The customers
need a reliable way to assure access to Area 0. Redundant routers with multiple links into Area
0 provide this reliability.
Note: Please note that these examples, and those in further sections, are given for your study and consideration only. They
are to help you reach a better understanding of the fundamental concepts before configuring your own application. It will
be necessary for you to modify these examples to match your own network design.
ProCurve 7203dl
Hostname: SecureRouter1
Ethernet 0/1
10.0.3.3/24
10.0.1.1/24
10.1.1.0/24
ProCurve 7203dl
Hostname: ABR
AREA 0
10.0.2.1/24
Ethernet 0/1
10.0.3.2/24
ProCurve 7102dl
Hostname: SecureRouter2
AREA 101
Figure 27: OSPF Solution Example 1
ABR (Area 0)
!
!
hostname "abr"
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
interface eth 0/1
ip address dhcp
70
no shutdown
!
interface eth 0/2
no ip address
shutdown
!
!
!
interface t1 3/1
speed 64
no shutdown
!
interface t1 3/2
speed 64
no shutdown
!
interface t1 3/3
shutdown
!
interface t1 3/4
shutdown
!
interface t1 3/5
shutdown
!
interface t1 3/6
shutdown
!
interface t1 3/7
shutdown
!
interface t1 3/8
shutdown
!
interface ppp 1
ip address 10.0.1.1
255.255.255.252
ip ospf network point-to-point
no shutdown
bind 1 t1 3/1 1 ppp 1

!
interface ppp 2
ip address 10.0.2.1
255.255.255.252
no shutdown
bind 2 t1 3/2 2 ppp 2
!
!
router ospf
network 10.1.1.0 0.0.0.255 area
0
network 10.0.1.1 0.0.0.255 area
101
network 10.0.2.1 0.0.0.255 area
101
area 0 default-cost 1
!
!
!
no ip tftp server
no ip http server
no ip snmp agent
no ip ftp agent
!
!
!
!
!
!
!
line con 0
no login
!
line telnet 0 4
login
password pnb
!
end
Secure Router 1
!
!
hostname "SecureRouter1"
no enable password
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
71
!
!
!
!
!
!
!
!
!
!
!
!
interface eth 0/1
ip address 10.0.3.3
255.255.255.0
no shutdown
!
interface eth 0/2
ip address dhcp hostname
"ProCurveSR7102dl"
shutdown
!
!
!
interface t1 1/1
speed 64
no shutdown
!
interface t1 1/2
line-length 0
shutdown
!
interface bri 1/3
shutdown
!
interface ppp 1
ip address 10.0.1.2
255.255.255.252
ip ospf network point-to-point
no shutdown
bind 1 t1 1/1 1 ppp 1
!
!
router ospf
network 10.0.3.1 0.0.0.255 area
101
network 10.0.1.2 0.0.0.255 area
101
area 101 default-cost 1
!
!
!
no ip tftp server
no ip http server
no ip snmp agent
no ip ftp agent
!
!
!
!
!
!
!
line con 0
no login
!
line telnet 0 4
login
!
end
Secure Router 2
!
!
hostname "SecureRouter2"
no enable password
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email

!
!
!
interface eth 0/1
ip address 10.0.3.2
255.255.255.0
no shutdown
!
interface eth 0/2
no ip address
shutdown
72
!
!
!
interface t1 3/1
speed 64
no shutdown
!
interface t1 3/2
shutdown
!
interface t1 3/3
shutdown
!
interface t1 3/4
shutdown
!
interface t1 3/5
shutdown
!
interface t1 3/6
shutdown
!
interface t1 3/7
shutdown
!
interface t1 3/8
shutdown
!
interface ppp 1
ip address 10.0.2.2
255.255.255.252
no shutdown
bind 1 t1 3/1 1 ppp 1
!
!
router ospf
network 10.0.3.1 0.0.0.255 area
101
network 10.0.2.1 0.0.0.255 area
101
!
!
!
no ip tftp server
no ip http server
no ip snmp agent
no ip ftp agent
!
!
!
!
!
!
!
line con 0
no login
!
line telnet 0 4
login
!
end
OSPF Solution Example 2

The customer has grown their business. Some new sectors were purchased, other sectors of
their business have grown internally. They would like each entity to administer their own
networks at each location as a separate entity or area. Some locations have RIP in their AS
so that must be integrated also.
73
OSPF
Stub Area 5
10.1.6.5
OSPF
Base Area 0
RIP
Secure Router
ASBR
10.1.1
1
11.1.1
12.1.9.2
10.1.5.2
10.1.3.2
2
Cisco 2621
10.1.2.1
10.1.8.3
Secure Router
Other
Secure
Router
ABR
10.1.5.5 Secure
Router
Stub
10.1.4.2
OSPF
Stub 4
10.1.4.4
10.1.7.4
Secure Router
Other Area
Figure 28: OSPF Solution Example 2
Base-Area.cfg
!
!
hostname "Other Area"
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
!
interface eth 0/1
ip address 10.1.7.4
255.255.255.0
no shutdown
!
interface eth 0/2
no ip address
shutdown

!
!
!
interface t1 1/1
speed 64
no shutdown
!
interface ppp 1
ip address 10.1.4.4
255.255.255.0
no shutdown
bind 1 t1 1/1 1 ppp 1
!
!
router ospf
network 10.1.4.0 0.0.0.255 area
4
!
!
!
no ip tftp server
no ip http server
74
no ip snmp agent
no ip ftp agent
!
!
!
line con 0
no login
!
line telnet 0 4
login
!
end
Other-Base-Area.cfg
!
!
hostname "Other Base Area"
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
!
interface eth 0/1
ip address 10.1.1.3
255.255.255.0
no shutdown
!
interface eth 0/2
ip address 10.1.8.3
255.255.255.0
no shutdown
!
!
!
interface e1 1/1
no shutdown
!
interface e1 1/2
no shutdown
!
!
router ospf
network 10.1.1.0 0.0.0.255 area
0
!
!
!
no ip tftp server
no ip http server
no ip snmp agent
no ip ftp agent
!
!
!
line con 0
no login
!
line telnet 0 4
login
password pnb
!
end
ABR.cfg
!
hostname "ABR"
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
75
!
!
!
interface eth 0/1
ip address 10.1.1.2
255.255.255.0
no shutdown
!
interface eth 0/2
ip address 10.1.3.2
255.255.255.0
no shutdown
!
!
!
interface t1 1/1
speed 64
no shutdown
!
interface t1 1/2
speed 64
no shutdown
!
interface ppp 1
ip address 10.1.4.2
255.255.255.0
no shutdown
bind 1 t1 1/1 1 ppp 1
!
interface ppp 2
ip address 10.1.5.2
255.255.255.0
lldp send system-capabilities
no shutdown
bind 2 t1 1/2 2 ppp 2
!
!
router ospf
network 10.1.1.0 0.0.0.255 area
0
network 10.1.4.0 0.0.0.255 area
4
network 10.1.5.0 0.0.0.255 area
5
area 5 stub no-summary
!
!
!
no ip tftp server
no ip http server
no ip snmp agent
no ip ftp agent
!
!
!
line con 0
no login
!
line telnet 0 4
login
password pnb
!
end
ASBR.cfg
!
!
hostname "ASBR"
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
!
76
!
!
!
interface eth 0/1
ip address 10.1.1.1
255.255.255.0
no shutdown
!
interface eth 0/2
ip address 10.1.2.1
255.255.255.0
no shutdown
!
!
!
interface t1 1/1
speed 64
no shutdown
!
interface t1 1/2
shutdown
!
interface t1 3/1
shutdown
!
interface t1 3/2
shutdown
!
interface t1 3/3
shutdown
!
interface t1 3/4
shutdown
!
interface t1 3/5
shutdown
!
interface t1 3/6
shutdown
!
interface t1 3/7
shutdown
!
interface t1 3/8
shutdown
!
interface modem 1/3
shutdown
!
interface ppp 1
ip address 11.1.1.1
255.255.255.0
no shutdown
bind 1 t1 1/1 1 ppp 1
!
!
router rip
redistribute ospf
network 11.1.1.0 255.255.255.0
!
!
router ospf
network 10.1.1.0 0.0.0.255 area
0
redistribute rip
!
!
!
no ip tftp server
no ip http server
no ip snmp agent
no ip ftp agent
!
!
!
!
!
!
!
line con 0
no login
!
line telnet 0 4
login
password pnb
!
end
77
Stub.cfg
!
!
hostname "Stub"
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
!
!
!
!
!
!
!
!
!
!
!
interface eth 0/1
ip address 10.1.6.5
255.255.255.0
no shutdown
!
interface eth 0/2
no ip address
shutdown
!
!
!
interface t1 1/1

speed 64
no shutdown
!
interface t1 1/2
line-length 0
shutdown
!
interface ppp 1
ip address 10.1.5.5
255.255.255.0
no shutdown
bind 1 t1 1/1 1 ppp 1
!
!
router ospf
network 10.1.5.0 0.0.0.255 area
5
area 5 stub
!
!
!
no ip tftp server
no ip http server
no ip snmp agent
no ip ftp agent
!
!
!
!
!
!
!
line con 0
no login
!
line telnet 0 4
login
!
end
78
Cisco 2621xm.cfg
!
version 12.3
msec
service timestamps log datetime
msec
!
hostname 2621_left
!
boot-start-marker
boot-end-marker
!
!
no network-clock-participate slot
1
no network-clock-participate wic
0
no aaa new-model
ip subnet-zero
ip cef
!
!
!
!
!
!
!
!
!
no crypto isakmp enable
!
!
!
ip address 192.168.1.253
255.255.255.0
duplex auto
speed auto
!
interface Serial0/0
ip address 11.1.1.2
255.255.255.0

encapsulation ppp
backup delay 30 60
backup interface Dialer1
!
interface BRI0/0
255.255.255.0
shutdown
dialer string 7850254
dialer-group 1
!
ip address 12.1.9.2
255.255.255.0
duplex auto
speed auto
!
interface Dialer1
no ip address
!
router rip
network 11.0.0.0
!
ip classless
ip http server
!
!
dialer-list 1 protocol ip permit
!
!
control-plane
!
!
!
line con 0
line aux 0
line vty 0 4
password password
login
!
!
!
end
79
BGP Solution Example 1
ProCurve 7203dl
Hostname: ISPRouter1
Ethernet 0/1
10.3.0.1 /24
10.1.0.3 /24
10.1.0.1 /24
BGP AS 1
10.1.1.0 /24
10.2.0.3 /24
ProCurve 7203dl
Hostname: BGP3
BGP AS 3
10.2.0.2 /24
Ethernet 0/1
10.3.0.2 /24
ProCurve 7203dl
Hostname: ISPRouter2
BGP AS 2
Figure 29: BGP Solution Example 1
BGP3
!
!
hostname "bgp3"
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
interface eth 0/1
ip address 10.1.1.0 /24
no shutdown
!
interface eth 0/2
no ip address
shutdown
!
!
!
interface t1 3/1
speed 64
no shutdown
!
interface t1 3/2

speed 64
no shutdown
!
interface t1 3/3
shutdown
!
interface t1 3/4
shutdown
!
interface t1 3/5
shutdown
!
interface t1 3/6
shutdown
!
interface t1 3/7
shutdown
!
interface t1 3/8
shutdown
!
interface ppp 1
ip address 10.1.0.3
255.255.255.0
no shutdown
bind 1 t1 3/1 1 ppp 1
!
interface ppp 2
ip address 10.2.0.3
255.255.255.0
no shutdown
80
bind 2 t1 3/2 2 ppp 2

!
!
router ospf
network 10.1.1.0 0.0.0.255 area
0
!
router bgp 3
no auto-summary
no synchronization
bgp router-id 10.1.0.3
neighbor 10.1.0.1
remote-as 1
neighbor 10.2.0.2
remote-as 2
!
!
no ip tftp server
no ip http server
no ip snmp agent
no ip ftp agent
!
!
!
!
!
!
!
line con 0
no login
!
line telnet 0 4
login
password pnb
!
end
ISP Router 1
!
!
hostname "ISProuter1"
no enable password
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
interface eth 0/1
ip address 10.3.0.1
255.255.255.0
no shutdown
!
interface eth 0/2
ip address dhcp hostname
"ProCurveSR7102dl"
shutdown
!
!
!
interface t1 1/1
speed 64
no shutdown
!
interface t1 1/2
line-length 0
shutdown
!
interface t1 2/1
speed 64
no shutdown
!
interface t1 2/2
shutdown
!
interface bri 1/3
shutdown
!
interface modem 2/3
shutdown
!
interface ppp 1
ip address 10.1.0.1
255.255.255.0
no shutdown
81
bind 1 t1 1/1 1 ppp 1

!
interface ppp 2
ip address 10.4.0.1
255.255.255.0
no shutdown
bind 2 t1 2/1 2 ppp 2
!
!
!
router bgp 1
no auto-summary
no synchronization
255.255.255.0
255.255.255.0
255.255.255.0
neighbor 10.1.0.3
remote-as 3
neighbor 10.3.0.2
remote-as 2
!
!
no ip tftp server
no ip http server
no ip snmp agent
no ip ftp agent
!
!
!
!
!
!
!
line con 0
no login
!
line telnet 0 4
login
!
end
ISP Router 2
!
!
hostname "ISProuter2"
no enable password
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
interface eth 0/1
ip address 10.3.0.2
255.255.255.0
no shutdown
!
interface eth 0/2
no ip address
shutdown
!
!
!
interface t1 3/1
speed 64
no shutdown
!
interface t1 3/2
shutdown
!
interface t1 3/3
shutdown
!
interface t1 3/4
shutdown
!
interface t1 3/5
shutdown
!
interface t1 3/6
shutdown
!
interface t1 3/7
82
shutdown
!
interface t1 3/8
shutdown
!
interface ppp 1
ip address 10.2.0.2
255.255.255.0
no shutdown
bind 1 t1 3/1 1 ppp 1
!
!
!
router bgp 2
no auto-summary
no synchronization
255.255.255.0
255.255.255.0
neighbor 10.2.0.3
remote-as 3
neighbor 10.3.0.1
remote-as 1
!
!
no ip tftp server
no ip http server
no ip snmp agent
no ip ftp agent
!
!
!
line con 0
no login
!
line telnet 0 4
login
!
end
Customer Site 2
!
!
hostname "Customer2"
enable password pnb
!
ip subnet-zero
ip classless
ip routing
!
event-history on
no logging email
!
!
!
!
interface eth 0/1
no ip address
shutdown
!
interface eth 0/2
no ip address
shutdown
!
!
!
interface t1 3/1
speed 64
no shutdown
!
interface t1 3/2
shutdown
!
interface t1 3/3
shutdown
!
interface t1 3/4
shutdown
!
interface t1 3/5
shutdown
!
interface t1 3/6
shutdown
!
interface t1 3/7
shutdown
!
interface t1 3/8
shutdown
83
!
interface adsl 2/1
shutdown
!
interface ppp 1
ip address 10.4.0.2
255.255.255.252
no shutdown
bind 1 t1 3/1 1 ppp 1
!
!
!
ip route 0.0.0.0 0.0.0.0 10.4.0.1
!
no ip tftp server
no ip http server
no ip snmp agent
no ip ftp agent
!
!
!
line con 0
no login
!
line telnet 0 4
login
password pnb
!
end
Feature Set Consideration for ProCurve Secure Router 7000dl Series

The ProCurve Secure Router 7000dl series supports the following Layer 3 protocols as
of this writing:
o
o
o
o
Static Routing
RIP version 1 and 2
OSPF
BGP 4
RIP versions must be set to only version 1 or version 2. There is currently no

compatibility mode.
Check http://procurve.com for the latest free software update, or for additions to the
physical interfaces supported.
How to Use the ProCurve Secure Router 7000dl Series

Both the 7102dl and 7203dl work effectively with all protocols. The 7203dl has the
edge in memory; 7102dl at 128Mbytes and the 7203dl at 256Mbytes.
Additional Topics
The following topics are beyond the scope of this paper. Never the less they are covered here
as overview information so the designer can make better decisions regarding these topics and
be reminded that these are aspect they should consider after the basic design is met.
High Availability and Redundancy

The designer needs to keep in mind aspects of the customer need for resiliency in their network.
Redundancy needs to be considered at all layers of the OSI model.
Redundancy can be built into the customer network at Layer 1 and 2 through optional backup
modules, MLFR, MLPPP. Remember that ISDN will operate at up to 128K bps but is far short of
the full speed of a T1 or E1.
Layer 3 redundancy methods include designs that use multiple routes between sites. An
example of an OSPF configuration is included in this guide for consideration. For the client to
router redundancy, the use of VRRP on the routing switches at the distribution layer are
recommend so that if the primary router goes down, the clients still have connectivity through
one of the switches.
Often the customer would like to use a different carrier for redundant paths. The Internet can
provide such an option. The primary link can be through a private carrier and utilize PPP or
84
Frame Relay. The backup link would be to the ISP. Any interface can be used to gain access to
the Internet and route selection can be controlled through the use of administrative distance
parameters.
See the ProCurve Secure Router 7000dl Series documentation set for further information.
IP Multicast
From a branch office perspective there is little requirement for extensive routing of multicasts.
There is a requirement for many to forward multicasts through the router to the central site.
The ProCurve Secure Router 7000dl series use a multicast route helper to allow for this.
Use the ip mcast-stub helper-address command to specify an IP address toward which IGMP
host reports and leave messages are forwarded. This command is used in IP multicast stub
applications in conjunction with the ip mcast-stub downstream and ip mcast-stub upstream
commands. Use the no form of this command to return to default. If there are multiple hops
from the WAN edge (central office) to the remote site, GRE-tunneling can be used to from a
virtual path between the central site and the remote site. This tunnel can be used as the
foundation for multicast stub routing.
The helper-address is configured globally and applies to all multicast-stub downstream
interfaces. The address specified may be the next upstream hop or any upstream address on
the distribution tree for the multicast source, up to and including the multicast source. The
router selects, from the list of multicast-stub upstream interfaces, the interface on the shortest
path to the specified address. The router then proxies, on the selected upstream interface
(using an IGMP host function), any host joins/leaves received on the downstream interface(s).
The router retransmits these reports with addresses set as if the report originated from the
selected upstream interface.
For example, if the router receives multiple joins for a group, it will not send any extra joins out
the upstream interface. Also, if it receives a leave, it will not send a leave until it is certain that
there are no more subscribers on any downstream interface.
See the SROS Command Line Interface Reference Guide for further information.
Security, Access Control Lists, and Virtual Private Networks

Security, which include many more facets than simply firewalls and access policies, should be
considered during the design phase. This is especially critical if the routers will connect to a
public network such as the Internet.
See the SROS Command Line Interface Reference Guide and the Virtual Private Network
(VPN) Configuration Guide in the ProCurve Secure Router 7000dl Series documentation set for
further information.
Quality of Service
Quality of Service should be considered apart from the WAN alone. How the customer uses the
network will play heavily in these decisions. When building the converged network utilize the
solution design guides at the http://www.procurve.com web site.
See the Low-latency Queuing Configuration Guide, and the SROS Command Line Interface
Reference Guide in the ProCurve Secure Router 7000dl Series documentation set for further
information.
85
Appendix A Route summarization

Routing among locations
10.3.0.0/24
10.3.1.0/24
10.3.255.0/24
(up to 255 networks)
Intranet core
Networks 10.0.0.0/24
through 10.0.255.0/24
Location C
Hosts in address range:
10.3.0.0 10.3.255.255
10.2.0.0/24
10.2.1.0/24
10.2.255.0/24
10.1.0.0/24
10.1.1.0/24
10.1.255.0/24
Location B
10.2.0.0 10.2.255.255
Location A
10.1.0.0 10.1.255.255
Figure 30: Routing among Locations
In this example, the routing infrastructure supports more than 750 user networks distributed
across three physical locations. While the technologies in place are similar to earlier examples,
which showed only eight user networks, the complexity of this topology presents a few new
challenges.
For instance, because there are so many networks at each location, it is inefficientand
possibly impossibleto connect every router to the intranet core. Instead, the topology features
redundant links among routers at each location. Another layer of aggregates the traffic from the
hosts at each location and connects that router to the core. This multi-layered hierarchical
approach can be scaled to support a network with hundreds of locations, if necessary.
Dynamic route exchange

Intranet core
Networks:
10.0.0.0/24-10.0.255.0/24
10.0.0.0/24-10.0.255.0/24
10.2.0.0/24-10.2.255.0/24
10.3.0.0/24-10.3.255.0/24
10.0.0.0/24-10.0.255.0/24
10.1.0.0/24-10.1.255.0/24
10.2.0.0/24-10.2.255.0/24
10.3.0.0/24-10.3.255.0/24
10.0.0.0/24-10.0.255.0/24
10.1.0.0/24-10.1.255.0/24
10.3.0.0/24-10.3.255.0/24
10.1.0.0/24-10.1.255.0/24
Location A
10.1.0.0 10.1.255.255
Location C
10.3.0.0 10.3.255.255
10.2.0.0/24-10.2.255.0/24
Location B
10.2.0.0 10.2.255.255
Each router may

have up to 1,024
route table
entries
Figure 31: Dynamic Route Exchange
A topology such as the one shown can result in very inefficient route tables. In the example,
each location includes 255 networks. Consequently, if every router advertises every route table
entry to all of its neighbors, every router in the entire intranet will have over 750 entries in its
86
route table. This is highly inefficient because it is not necessary for every router to know every
network. The number of connections between each location and the intranet core is limited to
one or two. Consequently, for each location all traffic destined for networks at other locations
will travel on one link or, if the routers support ECMP, will be shared by both links.
To avoid this inefficiency, IP network designers usually assign contiguous address space to
physically separated locations, regardless of whether they are buildings within the same campus
separated by a short distance or campuses within a larger enterprise that are separated by a
greater distance. This makes it possible to summarize the address space, enabling a larger
number of physical networks to be represented by a single route table entry.
Network summarization
Often, routers at a location have a limited number of paths to the networks within a given
address range. In these cases, you can increase routing efficiency by replacing many individual,
specific network advertisements with a single statement that specifies a larger range of
addresses using a shorter mask. In all cases, a shorter mask specifies a larger address range
and a longer mask specifies a shorter range. Any starting address with a 24-bit mask specifies a
range with 256 addresses. A starting address with a 16-bit mask specifies a range of 65,536
addresses.
This process is known as network summarization. In most vendor implementations, neither
RIP nor OSPF performs this summarization automatically; both require that you perform some
additional configuration steps to enable network summarization.
For example, in a hierarchical network, all traffic from Location A that is destined for other
locations must go through one of the routers connected to the core.
Summarization of address space using static routes

Intranet core router is configured with two static
summary routes toward address range 10.1.0.0/16
Intranet core
10.0.0.0/16
One specifies A1 as next hop

Another specifies A2 as next hop
A2
Routers A1 and A2:
Location A
10.1.0.0/16
A1
Are configured with static

default route 0.0.0.0/0;
next hop is a core router
interface
Include the static route in
RIP updates they send to
the edge routers in
Location A
A14
A13
A11
A12
Figure 32: Summarization of address space using static routes
In networks that implement RIP, static routes usually provide the mechanism for network
summarization.
In this example, network summarization will prevent routers in Location A from obtaining
detailed, specific advertisements for every network in the intranet. This process requires two
steps:
1. Disable the operation of RIP on both sides of the links that connect Routers A1 and A2
to the intranet core. This, of course, prevents the routers in Location A from processing
RIP advertisements sent from the core.
2. Define static routes for the path to networks or address ranges that do not appear as
more specific routes in the route able.
87
In the example, the goal is to provide a path for hosts at Location A to reach all destinations
that cannot be found at Location A, including addresses on the public Internet. To accomplish
this, you would specify the default route (0.0.0.0/0) that uses an intranet core interface as the
next hop.
While the core router may use a default static route to reach addresses in the pubic Internet, it
cant use the default route to reach hosts at different locations. Instead, the intranet core might
have a summarized route to each location. Because the addressing scheme is hierarchical, and
all hosts between 10.1.0.0 and 10.1.255.255 are at Location A, you can define a static
summary route for the path the core router has to Location A with the starting address 10.1.0.0
and a 16-bit mask (10.1.0.0/16). The 16-bit mask defines a range of over 65,536 addresses,
although some number of the addresses in this range would be inappropriate for host
addressing purposes.
Because the mask is 16 bits long, the first two octets of a destination address must match with
the starting address (10.1.0.0). The values in the remaining 16 bits are irrelevant for the
purposes of determining whether a given address is within the range. Any combination of zeros
and ones in the last 16 bits constitutes an unconditional match.
This means that every address that starts with a 10 in the first octet, has a 1 in the second
octet, and has any binary combination between 00000000 00000000 (0.0.) and 11111111
11111111 (255.255) is a member of the range specified by the starting address 10.1.0.0 and
the mask of /16. This includes all addresses between 10.1.0.0 and 10.1.255.255.
Although the diagram shows detailed operation only for Location A, the same procedures would
be used for other locations. The intranet core router(s) would need to have static routes
specifying each of the locations address ranges. It would forward traffic destined for a given
address range in the direction of the appropriate location. The routers that connect each
location to the intranet core will use the default route to forward all traffic for which it does not
have a more specific route in its route table.
Many routers treat the default static route as a special case of static route. That is, without
special configuration, some routers will not place the default route into its route table, even if it
advertised within a RIP update. Typically, if a router does not automatically listen for or accept
the default route, it is usually possible to selectively enable default route listening or to enable it
for all RIP interfaces on the router.
Summarizing remote address space

Without network summarization, a large enterprise with many locations connected through a
limited number of links to a centralized core may create an unnecessary burden on routers.
Memory and route table space are used inefficiently when the route table on a router describes
specific routes to every remote network and all of them have the same next hop.
RIP networks offer several options for summarizing address space. In one approach, you can
define each location with a separate classful network address range and have the routers
summarize all of the specific routes into the classful network.
For example, if you have 16 or fewer locations, you could assign Class B public addresses to
each location and configure the routers that connect the location to the core to summarize all of
the specific routes into the classful network. The Class B public address range is 172.16.0.0/12,
which means that there are 16 individual Class B networks: 172.16.0.0/16, 172.17.0.0/16,
172.18.0.0/16, up to 172.31.0.0/16. Each location could support 65,536 addresses, which
might be divided into 256 subnets, each of which has a 24-bit mask. Networks at Location 1, for
example, might be 172.16.0.0/24, 172.16.1.0/24, 172.16.2.0/24, 172.16.3.0/24 and so on.
Other locations would be configured similarly.
The routers that connect the location to the core would be configured to automatically
summarize all of the specific networks into the Class C parent network such as 172.16.0.0/16.
88
Summarizing at classful network boundaries
Intranet core
Networks 172.16.0.0 /24
- 172.16.255.0 /24
Advertise
172.16.0.0. /16
172.18.0.0. /16
172.19.0.0. /16
Advertise
172.16.0.0. /16
172.18.0.0. /16
172.19.0.0. /16
Advertise 172.17.0.0 /16
Location A
Networks 172.17.0.0. 24
172.17.255.0 /24
Advertise
172.16.0.0. /16
172.18.0.0. /16
172.19.0.0. /16
Advertise 172.19.0.0 /16
Location C
Networks 172.19.0.0. 24
172.19.255.0 /24
Advertise 172.18.0.0 /16
Location B
Networks 172.18.0.0. 24
172.18.255.0 /24
Figure 33: Example of Classful Network Summarization
In this example, the administrator of an intranet using Class B private addresses for the
enterprise has allocated network numbers along classful boundaries. All of the networks at a
given location are within the same Class B network.
If each location is connected to the core by 5300xl/3400cl routing switches, the switches will
automatically summarize the subnets at classful boundaries. All such networks will be
summarized if they do not extend into the network that connects the location to the core.
For example, the link that connects Location A to the core must not be within the address range
172.17.0.0/16. An appropriate address for this network (to enable auto-summarization) would
be within the 172.16.0.0/16 range, the range assigned to the core. You could assign to this
network a Class C private address such as 192.168.10.0/24. Basically, any address that is not
within 172.17.0.0/16 would allow auto-summarization to work.
However, if you assigned individual class C private networks to each of the location-to-core
links, each of these networks would appear in every routers route table at all locations. These
network addresses could not be automatically summarized. They are not subnets of a larger
parent network, but, from a classful perspective, each network in the range 192.168.0.0/24192.168.255.0/24 is a separate network.
It would be more efficient to make the location-to-core links a part of the range assigned to the
core. That way they would be part of the larger summary network and not represented
individually in each route table (across the entire intranet).
Summarizing a larger address space

Static routes are completely flexible. As long as an address range can be summarized using a
starting address and mask, you can create a static route for it. In the example above, an
administrator has used a single statement to define a range that includes all 16 private Class B
networks. Consequently, each route table requires one static route instead of three.
Note that the specific networks 172.17.10.0/24 and 172.17.30.0/24 are contained within the
summarized address range 172.16.0.0/12. Actually, that range contains 1,048,576 (65,536 *
16) addresses.
When the router receives packets to be forwarded, it runs a matching operation against route
table entries and forwards the traffic to the next hop (gateway) associated with the most
specific match.
89
In the example, a packet with the destination address 172.17.10.10 matches the first and
second entries. The second entry, 172.17.10.0/24, is more specific because it has a longer
mask. The router will forward the packet accordingly.
The summarization rules provide a useful guideline for address range assignment. Suppose, for
instance, that you were developing the IP addressing scheme for the Northwest campus of
ProCurve University, which uses 10.2 for the first octets. The lowest numbered range of four
networks would have starting addresses of 10.2.0.0 and mask of 255.255.252.0 or 22 bits.
The four networks in the range are:
10.2.0.0/24
10.2.1.0/24
10.2.2.0/24
0.2.3.0/24
Other ranges of four networks are expressed as 10.2.4.0/22, 10.2.8.0/22, 10.2.12.0/22,

10.2.16.0/24 etc. all the way up to 10.2.240.0/22, 10.2.244.0/24, 10.2.248.0/24, and
10.2.252.0/24.
The complete set of starting addresses for a range of eight networks that each have a 24-bit
mask are 10.2.0.0/21, 10.2.8.0/21, 10.2.16.0/24, 10.2.24/0/24, all the way up to
10.2.224.0/24, 10.2.232.0/21, 10.2.240.0/21, and 10.2.248.0/21. Since all of the addresses in
the address scheme that have a 2 in the second octet are reserved for the Northwest campus,
the entire third octet can be distributed among the networks at this location.
Appendix B Glossary
The following glossary is included for convenience and is not exhaustive. This paper focuses on
WAN design and so this glossary. For further information the designer should consult the many
references listed on the Internet along with the references noted at the end of this paper.
2B1Q
Short for 2 Binary, 1 Quaternary. 2B1Q is a full-duplex digital signaling technique used by many
digital communications technologies (like ISDN) to send data over a single pair of wires. It uses
a system of three different voltages: one for each of the two binary states (the 2B part of
"2B1Q") and a third, quaternary voltage that indicates both ends of the data connection are
sending the same binary value.
AAL
ATM Adaptation Layer The standards layer that allows multiple applications to have data
converted to and from the ATM cell. A protocol used that translates higher layer services into
the size and format of an ATM cell.
ADPCM
Adaptive Differential Pulse Code Modulation A technique for converting sound or analog
information to binary information (a string of 0s and 1s) by taking frequent samples of the
sound and expressing the value of the sampled sound modulation in binary terms.
ADSL
Asymmetric Digital Subscriber Line Supports two way asymmetrical transmission of data over
voice lines, downstream data bandwidth differs from upstream data bandwidth. Check current
implementation for data rates.
AMI
Alternate Mark Inversion Line-coding used with a T1 circuit where zeros are transmitted as
zeros and ones are transmitted as pulses with alternating polarity.
Analog
Continuously varying electrical signal in the shape of a wave used for voice or data
transmission.
ANI
Automatic Number Identification A service that provides the receiver of a telephone call with
the number of the calling phone.
90
ANSI
American National Standards Institute The primary organization for fostering the development
of technology standards in the United States.
AO/DI
Always On/Dynamic ISDN Allows the BRI D channel to be used for low-speed data
connection.
ASCII
American Standard Code for Information Interexchange ASCII is the most common format for
text files in computers and on the Internet. In an ASCII file, each alphabetic, numeric, or special
character is represented with a 7-digit binary number (a string of seven 0s or 1s). 128 possible
characters are defined.
ASIC
Application-Specific Integrated Circuit A silicon chip designed for a special application, such as
a particular kind of transmission protocol or hand-held computer.
Asymmetric Connection
A connection where data can flow in one direction at a much higher speed than in the other.
Some examples of asymmetric connections are ADSL, 56K Modems, and satellite downlinks.
Asynchronous
A method of data transmission which allows characters to be sent at non-predetermined
intervals by preceding each character with a start bit and ending each character with a stop bit.
ATM
Asynchronous Transfer Mode A dedicated, packet switched technology that allows high-speed
transmission of data through the use of small, fixed-length packets (cells). These cells are of 53
bytes, with 48 bytes being payload and the remaining 5 bytes for header information. The
transfer mode in which the information is organized into cells. It is asynchronous in the sense
that the recurrence of cells containing information from an individual user is not necessarily
periodic.
Authentication
The process of establishing an information source or users identity for secure transactions like
virtual private networking.
B8ZS
Bipolar 8 Zero Substitution The coding used to maintain ones density on a T1 circuit. This is
done by the insertion of two deliberate bipolar violations (two consecutive ones having the same
voltage) which replace the customers consecutive zero bits.
BECN
Backward Explicit Congestion Notification A bit set by a Frame Relay network to notify an
interface device (DTE) that congestion avoidance procedures should be initiated by the sending
device.
BER
Bit Error Rate The percentage of bits that have errors relative to the total number of bits
received in a transmission usually expressed as ten to a negative power. For example, a
transmission might have a BER of 10 to the minus six, meaning that out of 1,000,000 bits
transmitted, one bit was in error.
BERT
Bit Error Rate Test A procedure or device that measures the BER for a given transmission.
BISDN
Broadband Integrated Services Digital Network Any circuit capable of transmitting more than
one Basic Rate ISDN.
BGP
91
Border Gateway Protocol A protocol for exchanging routing information between gateway hosts
(each with its own router) in a network of an autonomous system. BGP is often the protocol
used between gateway hosts on the Internet.
BOC
Bell Operating Company A term for any of the 22 original companies (or their successors) that
were created when AT&T was broken up in 1983 and given the right to provide local telephone
service in a given geographic area.
BONDING
Bandwidth ON Demand INteroperability Group This group develops the standard that provides
for the aggregation of data from multiple ISDN calls into a coherent data stream.
bps
Bits per second In data communications, a common measure of data speed for computer
modem and transmission carriers. As the term implies, the speed in Bps is equal to the number
of bits transmitted or received each second.
BRI
Basic Rate Interface ISDN A user to network interface consisting of two 64KBIT/s bearer (B)
channels and one 16KBIT/s signaling (D) channel. The B channels carry data, voice or video
traffic. The D channel is used to set up calls on the B channels and carry packet data.
Bursty/Burstiness
Sporadic use of bandwidth that does not use the total bandwidth of a circuit 100 percent of the
time.
Caller ID
Caller ID is a telephone company feature that notifies a telephone being called of who is (or at
least what phone number is) originating the call. On analog POTS phone systems, Caller ID
information is transmitted to the telephone set between the first and second ring of the phone.
On ISDN sets, Caller ID data is sent as part of the Q9.31 "call setup" information sent of the
ISDN D channel. Some states, like California, regulate the implementation of Caller ID very
strictly, requiring that phone companies offer their customers the option of keeping their
numbers private when placing a call.
Carrier
(1) Telecommunications term for the PTT, LEC, or RBOC that provides the physical medium and
the transmission technology for connecting customers to the resources allowing Wide Area
Networking. (2) An alternating-current wave of constant frequency, phase and amplitude. By
varying (modulating) the frequency, phase or amplitude of a carrier wave, information is
transmitted.
C-Bit Parity
Framing format for a DS-3 signal. This provides possibilities for in-service, end-to-end path
performance monitoring of the DS-3 signal, and in-band data links.
CCITT
Consultative Committee on International Telegraphy and Telephony The CCITT, now known as
the ITU-T (International Telecommunications Union Telecommunications Services Sector), is
the primary international body for fostering cooperative standards for telecommunications
equipment and systems.
CDMA
Code Division Multiple Access One of the three wireless telephone transmission technologies.
After digitizing the data, it spreads it out over the entire bandwidth it has available. Multiple
calls are overlaid over each other on the channel, with each assigned a unique sequence code.
CDR (Call Detail Record)
A data record typically used in a telephony system to record usage information on a per-call
basis. Typical fields in the record include originating number, terminating number, start-time,
duration, etc.
Content Delivery Network or Content Distribution Network
92
Intelligent or content-aware networks that enable dependable and responsible delivery of

content to the edge of the Internet or the eventual end user.
CDSL
Consumer Digital Subscriber Line A trademarked version of DSL that is somewhat slower than
ADSL (1 Mbps downstream, probably less upstream) but has the advantage that a splitter
does not need to be installed at the users end.
Channelized
A circuit that is created by the multiplexing-demultiplexing voice and/or data bandwidth using
analog or digital techniques.
CIR
Committed Information Rate The committed rate (usually less than the access rate) which the
carrier guarantees to be available to transfer information to its destination under normal
circumstances for a particular PVC.
CLEC
Competitive Local Exchange Carrier In the United States, a CLEC is a company that competes
with the already established local telephone business by providing its own network and
switching. The term distinguishes new or potential competitors from established local exchange
carriers and arises from the Telecommunications Act of 1996 which was intended to promote
competition among both long-distance and local phone service providers.
CMIP
Common Management Information Protocol A network management protocol built on the Open
Systems Interconnection (OSI) communication model. The related Common Management
Information Services (CMIS) defines services for accessing information about the network
objects or devices, controlling them, and receiving status reports from them.
CO
Central Office In the United States, an office in a locality to which subscriber home and
business lines are connected on a local loop. It has equipment that can switch calls locally or to
long-distance carrier phone offices.
CoS
Class of Service A way of managing traffic in a network by grouping similar types of traffic
together and treating each type as a class with its own level of service priority.
CPE
Customer Premises (Provided) Equipment Equipment that is on the end user's side of the
network interface and is not usually owned by the Local Exchange Carrier. Some examples of
this equipment are CSU/DSUs, modems and telephones.
CRC
Cyclic Redundancy Check A method of checking for errors in data that has been transmitted on
a communications link. A sending device applies a 16- or 32-bit polynomial to a block of data
that is to be transmitted and appends the resulting cyclic redundancy code to the block.
CRM
Customer Relationship Management An information industry term for methodologies, software,
and usually Internet capabilities that help an enterprise manage customer relationships in an
organized way.
CSP
Competitive Service Provider All companies in competition to deliver telecommunications
services to both businesses and individuals.
CSU
Channel Service Unit A device that provides an accessing arrangement at a user location to
either switched or pointto- point, data-conditioned circuits at a specifically established data
signaling rate. Note: A CSU provides local loop equalization, transient protection, isolation, and
Central Office loop-back testing capability.
CTI
93
Computer Telephony Integration The use of computers to manage telephone calls.

D4 (SF)
A T1 transmission Superframe format consisting of 12 frames of 192 bits each plus an additional
193rd bit used for link control and error checking.
DACS
Digital Access & Cross-connect System Allows T1 carrier facilities (or any other subchannels) to
be switched or cross-connected to another T1 carrier.
DBU
Dial Back-Up A means of providing disaster recovery for a circuit that has failed.
DCE
Data Communications Equipment A device which provides all the functions required for
connection to local exchange carrier's lines and for converting signals between telephone lines
and DTE.
DDS
Dataphone Digital Service or Digital Data System A non-switched (dedicated) digital service
network for data rates of up to 56,000 bits per second typically only seen in North America.
DE
Discard Eligible A user-set mark indicating that a frame may be discarded in preference to
other frames if congestion occurs, to maintain the committed quality of service within the
network.
DES
Data Encryption Standard DES is a published encryption algorithm which uses a 56-bit
symmetric key to encrypt data in 64-bit blocks.
DHCP
Dynamic Host Configuration Protocol A protocol that lets network administrators manage
centrally and automate the assignment of Internet Protocol (IP) addresses within an
organizations network.
Digital Certificates
A digital certificate is an electronic means to establish a users identity when establishing a
virtual private network connection. It is issued by a certification authority (CA) and contains
your name, a serial number, expiration dates, a copy of the certificate holder's public key (used
for encrypting and decrypting messages).
DLC
Digital Loop Carrier Equipment that bundles a number of individual phone line signals into a
single-multiplexing digital signal for local traffic between a telephone company central office and
a business complex or other outlying service area.
DLCI
Data Link Connection Identifier A 10-bit field within the address field which identifies the data
link and its service parameters.
DNIS
Dialed Number Identification Service A telephone service that provides for the receiver of a call
the number the caller dialed. It is a common feature of 800 and 900 lines. If there are multiple
800 or 900 lines to the same destination, DNIS tells which number was called.
DNS
Domain Name Server Computers that convert domain names to IP addresses.
DOCSIS
Data Over Cable Service Interface Specification A project for the North American cable industry
aimed at developing specifications for cable modems and associated equipment.
DS-0
94
Digital Signal at the Zero Level A bipolar signal transmitted at 64Kbps. Often referred to as an
individual channel on a T1 which consists of a total of 24 DS-0s or channels.
DS-1
Digital Signal at the First Level The combination of 24 DS-0s creating a bipolar signal that is
transmitted at 1.544 Mbps. Also called T1.
DS-3
Digital Signal at the Third Level Circuit that can carry 672 DS-0 channels and has a bipolar
signal transmitted at 44.736 Mbps.
DSL
Digital Subscriber Line A family of digital lines provided by CLECS and telephone companies to
local customers.
DSLAM
Digital Subscriber Line Access Multiplexer a network device, usually at a telephone company
Central Office, that receives signals from multiple customer DSL connections and puts the
signals on a high-speed backbone line using multiplexing techniques.
DSP
Digital Signal Processor A specialized computer chip that performs a variety of complex
operations on digitized signals. DSP works by clarifying or standardizing, the levels or states of
a digital signal. A DSP circuit is able to differentiate between human-made signals, which are
orderly, and noise which is inherently chaotic.
DSU
Data Service Unit A device used for interfacing Data Terminal Equipment (DTE) to the public
telephone network. (2) A type of short-haul, synchronous-data line driver, usually installed at a
user location that connects user synchronous equipment over a 4-wire circuit at a preset
transmission rate to a servicing Central Office. Note: This service can be for a point-to-point or
multipoint operation in a digital data network.
DSX-1
Digital Cross-Connect Often used for connecting devices at 1.544 megabits. Also known as a
short-haul DS-1 (655 feet). Often used with a PBX.
DTE
Data Terminal Equipment Equipment that sends and receives information. An example of a
DTE is a user's PC.
DWDM
Dense Wavelength Division Multiplexing A technology
that puts data from different sources together on an optical fiber with each signal carried on its
own separate light wavelength. Using DWDM, up to 80 (and theoretically more) separate
wavelengths or channel of data can be multiplexing into a lightstream transmitted on a single
optical fiber. In a system with each channel carrying 2.5 Gps, up to 200 billion bits can be
delivered a second by the optical fiber. DWDM is also sometimes called Wave Division
Multiplexing (WDM).
E1
Similar to T1, see T1, yet with 32 DS0 channels running at a 2.048Mbps rate. Other differences
from T1 include framing type and encoding. HDB3 is the most common encoding scheme and is
similar to AMI.
Echo Cancellation
A device that support this technique provides filtering of unwanted signals that are called
echoes This technique is typically used for voice applications over packet-based networks.
Encryption
Scrambling data in such a way that it can only be unscrambled through the application using a
special key.
Error Rate
95
A measure of the performance of a digital transmission system. It can be specified as a bit error
rate (the probability of error per bit transmitted), as a block error rate (the probability of one or
more errors in a specified-length block of bits), or in other forms such as percent error-free
seconds.
Errored Second
Any 1-sec interval containing at least one bit error.
ESF
Extended Superframe Format A T1 format consisting of 24 consecutive frames that use the
framing bit to provide maintenance and diagnostic functions.
ESP
Encapsulated Security Payload The ESP provides confidentiality for packets, which are the
message units that the Internet Protocol deals with and that the Internet transports, by
encrypting the payload data to be protected.
European Telecommunications Standards Institute (ETSI)
The European equivalent of ANSI (American National Standards Institute).
FDL
Facilities Data Link An out-of-band, 4 kbit management path, available on ESF T-1s.
FDM
Frequency Division Multiplexing A technique in which numerous signals are combined for
transmission on a single communications line or channel. Each signal is assigned a different
frequency (subchannel) within the main channel.
FECN
Forward Explicit Congestion Notification A bit set sent by a Frame Relay network to notify an
interface device (DTE) that congestion avoidance procedures should be initiated by the receiving
device.
Fiber Optics
An optical transmission medium that consists of thin, plastic (or glass) strands which reflect
light pulses within their interior along their length as a means to transmit large amounts of
data.
Firewall
A security device that establishes a barrier to contain designed network traffic within a specified
area by allowing or denying access.
FRAD
Frame Relay Access Device A generic name for a device that multiplexes and formats traffic for
entering a Frame Relay network.
Frame Relay
A Layer 2, Packet-Based, carrier-switching technology. It provides features and benefits of a
dedicated DDS or T1 network, but without the expense of multiple dedicated circuits. Frame
Relay is deployed over the same services used to deploy DDS and T1. In a Frame Relay
network, circuits are connected to a packet switch within the network that ensures packets are
routed to the correct location.
FT1 or FrT1
Fractional T-1 A portion of a T-1 circuit. A full T-1 circuit has a capacity of 1.544 Mbps
composed of twenty-four (24) 64 kbps channels. A customer may lease a portion of the full
circuit to effect cost-savings. A fractional T-1 can only be configured in increments of 64 kbps or
a certain number of channels.
FTTC
Fiber To The Curb Refers to the installation and use of fiber optic cable directly to the curbs
near homes or any
business environment as replacement for plain old telephone service.
96
FX
Foreign Exchange A telephone service that allows a user to have a number with an exchange
that is not the normal exchange for their geographic area. The service is provided by linking the
users normal Central Office with the foreign central office by a leased line.
FXO
Foreign Exchange Office Office side of a Foreign Exchange.
FXS
Foreign Exchange Station Station side of a Foreign Exchange.
GUI
Graphical User Interface A user interface that substitutes graphics for characters.
HDLC
High level Data Link Control A group of protocols or rules for transmitting data between
network points (sometimes called nodes). In HDLC, data is organized into a unit (called a
frame) and sent across a network to a destination that verifies its successful arrival.
HFC
Hybrid Fiber Coax A telecommunication technology in which fiber optic cable and coaxial cable
are used in different portions of a network to carry broadband content (such as video, data, and
voice). Using HFC, a local cable TV company installs fiber optic cable from the cable head-end
(distribution center) to serving nodes located close to business and residential users and from
these nodes uses coaxial cable to individual businesses and homes.
HDSL
High-bit-rate Digital Subscriber Line This employs a 2B1Q modulation technique across the
same type pairs traditionally encountered with metallic T1 delivery systems. This satisfies Telco
distance requirements without the use of repeaters. (Usually used in a campus environment.)
HDSL2
High-bit-rate Digital Subscriber Line version 2 (also known as g.SHDSL or SHDSL) Allows
service providers to deliver full T-1 and possibly E-1 over a single twisted pair.
HIPPI
High Performance Parallel Interface A method of delivering high-speed point-to-point data
between supercomputers or high-end workstations and peripherals.
HSSI
High Speed Serial Interface A serial interface operating up to 52 Mbps and up to 50 ft.
IAD
Integrated Access Device A device which supports voice, data and video streams over a single
high-speed connection.
IDF
Intermediate Distribution Frame A freestanding or wallmounted rack for managing and
interconnecting the telecommunications cable between end-user devices and a Main Distribution
Frame (MDF).
IDSL
ISDN Digital Subscriber Line (2B1Q ISDN without the dial-up capabilities). A symmetric DSL
service that can transmit and receive data up to 144 Kbps.
IKE
Internet Key Exchange A protocol whose purpose is to negotiate and provide authenticated
keys for security associations in a protected manner. Processes which implement this protocol
can be used for negotiating virtual private networks (VPNs) to a secure host or network.
IEC
Interexchange Carrier A long distance carrier.
ILEC
97
Incumbent Local Exchange Carrier A telephone company in the United States that was
providing local service in a specific geographic area when the Telecommunications Act of 1996
was enacted. ILECs include the former Bell operating companies which were grouped into
holding companies known collectively as the Regional Bell Operating Companies (RBOC) when
the Bell System was broken up by a 1983 consent decree.
IP
Internet Protocol The method or protocol by which data is sent from one computer to another
on the Internet. Each computer (known as a host) on the Internet has at least one IP address
that uniquely identifies it from other computers on the Internet.
IPSec
A developing standard for security at the network or packet processing layer of network
communication. Earlier security approaches have inserted security at the application layer of the
communications model. IPSec will be especially useful for implementing virtual private networks
and for remote user access through public connection to private networks.
ISDF
Integrated Service Development Framework A set of CCITT/ITU standards for digital
transmission over ordinary telephone copper wire as well as over other media. Home and
business users who install an ISDN adapter (in place of a modem) can see highly-graphic web
pages arriving very quickly (up to 128 Kbps).
ISDN
Integrated Services Digital Network An extension of the digital transmission and common
channel signaling concepts of the public telephone network to the CPE. This includes BRI and
PRI.
ISO
International Standards Organization Group responsible for the creation of a set of CCITT/ITU
standards for digital transmission over ordinary telephone copper wire as well as over other
media.
ISP
Internet Service Provider A company that provides individuals and/or businesses access to the
Internet and other related services such as web site building and virtual hosting. An ISP has the
equipment and the telecommunication line access required to have POP on the Internet for the
geographic area served.
ITU (International Telecommunications Union)
A United Nations organization that establishes standards for telecommunications devices, like
ISDN hardware, modems, and Fax machines. ITU standards include J.112, J.122, H.323, V.90,
X.25, and X.500.
IXC
IntereXchange Carrier A telephone company that provides connections between local
exchanges in different geographic areas. IXCs provide interlocal access and transport area
service as described in the Telecommunications Act of 1996.
Kbps
Kilobits per second (one thousand bits per second)- A measure of bandwidth (the amount of
data that can flow in a given time) on a data transmission medium. Higher bandwidths are more
conveniently expressed in Megabit per second (Mbps or millions of bits per second) and in
Gigabits per second (Gbps, or billions of bits per second).
Key
In cryptography, a key is a unique value that is applied to encrypt and/or decrypted data. The
length of a key generally determines how difficult it will be to decrypt the data. Both Public and
Private keys are available and can be used together known as asymmetric encryption. A system
for using just public keys is called a public key infrastructure (PKI).
L2TP
98
Layer 2 Tunneling Protocol An extension of the Pointto- Point Tunneling Protocol (PPTP) used
by an Internet Service Provider (ISP) to enable the operation of a Virtual Private Network (VPN)
over the Internet.
LAN
Local Area Network A group of computers and associated devices that share a common
communications line and typically share the resources of a single processor or server within a
small geographic area.
LATA
Local Access and Transport Area A term in the United States for a geographic area covered by
one or more local telephone companies, which are legally referred to as local exchange.
LDN
Listed Directory Number Your seven digit phone number hence, the number that would be
listed in the phone directory.
LEC
Local Exchange Carrier The term for a public telephone company in the U.S. that provides local
service. These can be either one of the Bell operating companies or an independent.
LS
Loop Start A method of signaling to a switch to start a call, under which an off-hook condition
closes a circuit and causes current to flow, informing the switch to provide dial tone. (Compare
to ground start).
Local Loop
This portion of the telecommunications network physically connects end users to the central
office network facilities and generally is dedicated to that particular user. Twisted pairs of
copper wire form the traditional medium of the telephone network local loop, although other
connections now are used in some cases.
MAN
Metropolitan Area Network A network that interconnects users with computer resources in a
geographic area or region larger than that covered by a large local area network, but smaller
than the area covered by a wide area network. It typically extends as far as 50-kilometers and
operates at speeds between 1 Mbps to 200Mbps.
Mbps
Million (Mega) bits per second The measure of bandwidth (the total information flow over a
given time) on a data transmission medium such as twisted-pair copper cable, coaxial cable, or
optical fiber line.
MIPS
Millions of Instructions Per Second A general measure of computing performance and, by
implication, the amount of work a larger computer can do. Generally, this refers to the number
of instructions that can be processed by the CPU in a given second.
MTBF
Mean Time Between Failure A measure of how reliable a hardware product or component is.
For most components, the measure is typically in thousands or even tens of thousands of hours
between failures.
MTU
Maximum Transmission Unit the largest unit of data that can be sent across a given medium.
Multiplexing
Transmission of multiple signals over a single channel.
NAT
Network Address Translation The translation of an IP address used with one network to a
different IP address that is known within another network. One network is designated as the
inside network and the other is the outside.
99
NEBS
Network Equipment Building Standards NEBS testing is required for vendors who wish to sell
equipment to the Regional Bell Operating Companies (RBOCs) and the Competitive Local
Exchange Carriers (CLECs). Level 3 testing is the most stringent level of testing.
Network
A generic term describing the tying of like things together. In telecommunications, it usually
refers to infrastructure that provides for user voice and data transmission.
NEXT
Near End Crosstalk An error condition that can occur when connectors are attached to twisted
pair cabling. NEXT is usually caused by crossed or crushed wire pairs. The error condition does
not require that the wires be crushed so much that the conductors inside become exposed. Two
conductors only need to be close enough so that the radiating signal from one of the wires can
interfere with the signal traveling on the other.
NOC
Network Operations Center A place from which a telecommunications network is supervised,
monitored, and maintained. Enterprises with large networks and large network service providers
such as GTE Internetworking typically have a network operations center.
NNI
Network to Network Interface The connection between two public service network providers.
NT-1
Network Termination type 1 A Basic Rate ISDN-only device that converts a service provider's
U-Interface to a customers S/T interface. It can be stand-alone or integrated into a terminal
adapter.
OC-N
Optical Carrier Level N These are fundamental transmission rates for SONET with N=1, 3, 9,
12, 18, 24, 36, or 48.
OCUDP
Office Channel Unit Data Port Provides signal conversion from any rate of a customer's access
line (i.e. T1) to a 56 or 64Kbps single DDS line. This is often used with a DDS or a SW56
DSU/CSU.
OPX
Off Premise Extension The ability to have local access to a remote site (dial an extension)
across a Wide Area Network link.
OSPF
Open Shortest Path First A router protocol used within larger autonomous system networks in
preference to the Routing Information Protocol, an older routing protocol that is installed in
many of todays corporate networks.
PBX
Private Branch eXchange A customer premises switch that connects the customer location with
the public telephone network.
PCM
Pulse Code Modulation A digital scheme for transmitting analog data. The signals in PCM are
binary; that is, there are only two possible states represented by logic 1 (high) and logic 0
(low).
PCS
Personal Communications Service A wireless phone service somewhat similar to cellular
telephone service, but emphasizing personal service and extended mobility. It is sometimes
referred to as digital cellular (although cellular systems can also be digital). Like cellular, PCS is
for mobile users and requires a number of antennas to blanket an area of coverage.
PKI
100
Public Key Infrastructure PKI enables users to privately exchange data through a public
infrastructure like the Internet, through the use of a public and a private key pair that is
obtained and shared through a trusted authority.
PON
Passive Optical Network A system that brings optical fiber cabling and signals all or most of the
way to the end user. Depending on where the PON terminates, the system can be described as
Fiber-To-The-Curb (FTTC), Fiber-To-The-Building (FTTB), or Fiber-To-The-Home (FTTH).
PoP (POP)
Point of Presence A physical layer within a LATA at which an inter-LATA carrier establishes
itself for the purpose of obtaining LATA access and to which the local exchange carrier provides
access services.
PoP3 (POP3)
The current version of the most common protocol for receiving e-mail on a TCP/IP network.
PORT
The physical connection between a device and a circuit. Its capacity determines the greatest
amount of data that can be transmitted at any given time.
POTS
Plain Old Telephone Service Analog, voice-only telephone service.
PPP
Point-to-Point Protocol A protocol for communication between two computers using a serial
interface, typically a personal computer connected by phone line to a server. This protocol is
typically used for Internet connections originating from a dial-up line and a high-speed modem.
PPTP
Point-to-Point Tunneling Protocol A protocol that allows corporations to extend their own
corporate network through private tunnels over the public Internet.
PRI
Primary Rate Interface-ISDN A user to network interface consisting of 23 64Kbps bearer (B)
channels and one 64Kbps signaling (D) channel earned over a 1.544Mb/s DS-1 circuit. The B
channels carry data, voice or video traffic. The D channel is used to set up calls on the B
channels and carry packet data.
PSTN
Public Switched Telephone Network The public network that provides switched digital/analog
services to a customer/end user.
PTT
A term describing one of the telephone companies for most of the world outside of North
America.
PVC
Permanent Virtual Circuit The logical connection between two nodes. Each PVC is assigned a
CIR. A PVC can burst up to the port speed. See also SVC.
QoS
Quality of Service The quality of the telephone service provided to any given individual or
business.
RAS
Remote Access Server A computer and associated software that allows users to dial in or gain
access to a network remotely.
RBOC
Regional Bell Operating Company A term describing one of the U.S. Regional telephone
companies (or their successors) that were created as a result of the breakup of American
Telephone and Telegraph Company (AT&T, known also as the Bell System) by a U.S. Federal
Court consent decree on December 31, 1983. The seven original regional Bell operating
101
companies were Ameritech, Bell Atlantic, BellSouth, NYNEX, Pacific Bell. Southwestern Bell, and
US WEST. Each of these companies owned at least two Bell operating companies. The BOCs
were given the right to provide local phone service while AT&T was allowed to retain its long
distance service. The RBOCs and their constituent BOCs are part of the class of local exchange
carriers.
RBS
Robbed-Bit Signaling The process where the least significant bit in the 6th and 12th frame (of
a SF T1) and the 18 & 24th frame (of an ESF T1) is robbed for voice A, B, C, and D signaling
bits. These signaling bits indicate on/off hook conditions etc.
REN
Ringer Equivalency Number FCC Certification number approving a terminal telephone product
for direct sale to an end user as to not harm the network. The total number of RENs on one
telephone line must not exceed is 5.
RFC
Request For Comment The core method of specification for the Internet. (Some RFCs are
standards for the Internet.)
RIP
Routing Information Protocol A widely-used protocol for managing router information within a
self-contained network such as a LAN or an interconnected group of such LANs.
RJ45
A modular 8-wire jack/connector used with copper cable having four twisted pairs.
Router
Communications equipment which forwards information on a connectionless basis. More
specifically, a computer that connects different networks (LANs and WANs) at the Layer 3 of the
OSI Reference Model level often using IP addresses.
RSVP
Resource Reservation Protocol A protocol that allows channels or paths on the Internet to be
reserved for the multicast (one source to many receivers) transmission of video and other highbandwidth messages. RSVP is part of the Internet Integrated Service model, which ensures
best-effort service, real-time service, and controlled link sharing.
S/T Interface
A common way of referring to either an S or T Interface. This can be used to connect directly to
an ISDN 2B+D NT1 or an NT2 device with a terminal adapter.
SDH
Synchronous Digital Hierarchy A standard technology for synchronous data transmission on
optical media. It is the international equivalent of Synchronous Optical Network. Both
technologies provide faster and less expensive network interconnection than traditional
Plesiochronous Digital Hierarchy equipment.
SDLC
Synchronous Data Link Control The exclusive transport for a SNA network.
SDSL
Symmetric Digital Subscriber Line A DSL based on ISDN with 2B1Q, but is symmetric in nature
where both downstream and upstream traffic bandwidth may be up to 2.3 Mbps.
SHDSL
Symmetric High Bit Rate Digital Subscriber Loop SHDSL is defined by the new ITU Global
Standard G991.2 from February 2001 and provides high symmetric data rates with guaranteed
bandwidth and low interference with other telecommunications services.
SLA
Service Level Agreement A contract set up between an end-user and a service provider
outlining the guarantees of what will be provided for that digital service.
102
SONET
Synchronous Optical NETwork The U.S. (American National Standards Institute) standard for
synchronous data transmission on optical media. The international equivalent of SONET is
synchronous digital hierarchy (SDH). Together, they ensure standards so that digital networks
can interconnect internationally and that existing conventional transmission systems can take
advantage of optical media through tributary attachments.
SPID
Used only in North America, a Service Profile Identifier is a unique identifier that is
used to represent the service and feature identifiers of a particular Basic Rate ISDN
line or service provider. (This number generally is 10+ digits long and includes the LDN.)
The telephone company should give you your SPIDs at the time they assign you your ISDN
directory numbers.
SSL
Secure Sockets Layer A commonly used protocol for managing the security of message
transmission on the Internet. SSL uses a program layer located between the Internets
Hypertext Transfer Protocol (HTTP) and Transport Control Protocol (TCP) layers.
Stateful Inspection
Stateful Inspection is an advanced packet-filtering technology within firewalls that checks and
verifies each network packet in order to detect suspicious activity. Many network security
experts recommend the Stateful Inspection as the most trusted firewall technology.
STP
Shielded Twisted Pair A special kind of copper telephone wiring used in some business
installations. An outer covering or shield is added to the ordinary twisted pair telephone wires
and functions as a ground.
SVC
A and SVC, Switched Virtual Circuit, allows an end-user to establish on-demand data
connections between two end points through an ATM, Frame Relay, or X.25 network. See also
PVC.
Synchronous Transmission
A method of data transmission which allows characters to be sent in a continuous stream; the
beginning of one character contiguous with the end of the preceding one. Separation of
characters requires the receiver to maintain synchronization to a master timing source. It does
not use start and stop bits at the beginning and end of each byte to synchronize the data time
clocks at each end of a connection. Instead it sets its timing signal at the beginning and end of
each connection, and corrects discrepancies that arise over time by using the changing values
each device on the connection sends and receives to keep their clocks "in sync." Eliminating the
start and stop bits reduces the "overhead" required to transmit each byte, and allows for
increased throughput.
T1
A digital carrier signal designed to carry speech or data at the DS-1 rate. (See DS-1.)
T3
A digital carrier signal designed to carry speech or data at the DS-3 rate. (See DS-3.)
TMN
Telecommunications Management Network A network management model originated formally
in 1988 under the auspices of the International Telecommunication Union (ITU-TS) This was
done as a strategic goal to create or identify standard interfaces that would allow a network to
be managed consistently across all network element suppliers.
TDM
Time Division Multiplex Once digitized, voice/data signals from many sources may be combined
or multiplexed and transmitted over a single digital link.
TE1
Terminal Equipment type 1 Equipment that can directly connect to the ISDN line (often using
an S/ T Interface). Examples are ISDN phones, ISDN routers, ISDN
103
computers etc.
TE2
Terminal Equipment type 2 Equipment that is non- ISDN equipment. Needs an external
terminal adapter. Examples are PCs with EIA 232 interfaces and analog telephone sets.
Telco
American slang for Telephone Company The local telephone company.
Telnet
TCP/IP application that provides a terminal interface
between computers attached to a TCP/IP network. This allows a user at one site to interact with
a terminal at another site as if by a local terminal.
Trunk
An analog or digital connection from a circuit switch which carries user media content and may
carry telephony signaling (MF, R2, etc.). Digital trunks may be transported and may appear at
the Media Gateway as channels within a framed bit stream. Trunks are typically provisioned in
groups, each member of which provides equivalent routing and service.
Trunking
Transporting signals from one point (an antenna site for instance) to another point (such as a
headend), usually without serving customers directly. Trunking can be accomplished using
coaxial cable, fiber optics or microwave radio.
Tunnel
An established network connection in which data is encrypted and encapsulated for transmission
across a public or untrusted network, for eventual de-encapsulation and decryption.
U loop (or) U-Interface
An ISDN 2-wire digital circuit between the customer's network termination and the Local
Exchange Carrier's termination in the Central Office.
UNI
User to Network Interface Describes the connection between the user and the public network
service provider.
V.35
High Speed Digital Interface used for synchronous data rates up to approximately full Mbps
speed.
VAD
Voice Activation Detection A software application that allows a data network carrying voice
traffic over the Internet to detect the absence of audio and conserve bandwidth by preventing
the transmission of silent packets over the network. Most conversations include about 50
percent silence. VAD (also called silence suppression) can be enabled to monitor signals for
voice activity so that when silence is detected for a specified amount of time, the application
informs the Packet Voice Protocol and prevents the encoder output from being transported
across the network.
VC
Virtual Circuit A circuit or path between points in a
network that appears to be a discrete, physical path but is actually a managed pool of circuit
resources from which specific circuits are allocated as needed to meet traffic requirements.
Video Codec
A device that converts an analog signal into digital for transport and converts a digital signal
into analog for display.
VoATM
Voice over ATM A generic term for describing the delivery of voice services using Asynchronous
Transfer Mode.
104
VoDSL
Voice over Digital Subscriber Line A term used to describe the delivery of voice services using
Asynchronous Transfer Mode and DSL. The typical voice over DSL model includes an IAD for
delivering voice and data services to the end customer, a DSLAM for aggregating multiple DSL
lines into a single ATM cell stream and a voice gateway for interfacing back into the existing
public switched telephone network.
VoIP
Voice over IP A term used in IP telephony for a set of facilities for managing the delivery of
voice information using the Internet Protocol (IP). In general, this means sending voice
information in digital form in discrete packet rather than in the traditional circuit-committed
protocols of the public switched telephone. A major advantage of VoIP and Internet telephony is
that it avoids the tolls charged by ordinary telephone service.
VPN
Virtual Private Network A private data network that makes use of the public telecommunication
infrastructure, maintaining privacy through the use of tunneling protocol and security
procedures. A virtual private network can be contrasted with a system of owned or leased lines
that can only be used by one company.
WAN
Wide Area Network A high-speed network within a wide geographical area (usually larger than
a city or a metropolitan area) that shares data, programs, or equipment.
WDM
Wavelength Division Multiplexing A technology that puts data from different sources together
on an optical fiber, with each signal carried on its own separate light wave. Using WDM, up to
80 (and theoretically more) separate wavelengths or channel of data can be multiplexing into a
lightstream transmitted on a single optical fiber. In a system with each channel carrying 2.5
Gbps (One thousand million bits per second). Up to 200 billion bits can be delivered a second by
the optical fiber.
X.21
X.21 is a physical and electrical interface that uses two types of circuits: balanced (X.27N.1 1)
and unbalanced (X.26N.10). CCITT X.21 calls out the DB-15 connector. The physical interface
between the DTE and the local PTT-supplied DCE is defined in ITU-T recommendation X.21. The
DCE provides a full-duplex, bit-serial, synchronous transmission path between the DTE and the
local PSE. It can operate at data rates from 600bps to 64Kbps.
Appendix C Differences with Cisco Routers

Often entities that have been around the longest have made early decisions that subsequent
decisions, out of necessity, have been built upon. This is no different for any country,
company, or system. For example: carrier systems in North America have been in place for
many decades but have also decided to constrain newer technologies and demands for
bandwidth within those older systems; this is not news to anyone. Parts of the world where
there was not the same infrastructure have benefited by the ability to build new without
working on top of the old. The same goes with routers and their methods of configuration and
philosophies for structuring their architecture. The bottom line is that all systems will have
some differences due to these factors, and these factors alone should never reflect poorly on
any previous system or philosophy regarding the architecture decisions engineers and scientists
have had to make during a previous time.
Cisco routers have been available for building WAN networks for many years. Historical
decisions and their usage, it seems, have influenced subsequent features and their associated
commands. It is acknowledged that they have made the best engineering decisions they could
at that time, that required they accommodate their historical usage and the customer
requirements. As technology progressed so have these routers, yet many of the required
decisions for configuration appear to have historical usage influencing newer features and their
associated commands. This is not a poor reflection upon their decisions because this occurs for
any country, company, or system.
To help the designer and implementation engineer, this paper includes a table of major known
differences either with the command line interface configuration, or regarding a particular
philosophy for usage. It is not intended to show a best method but only allow the designer
105
and installer to readily discern the differences, and study the appropriate text to understand
more. The reader can also use the configuration examples in this guide to see some of these
differences.
This paper acknowledges that their will be future revisions to both router operating systems and
at some point the information contained below may not be completely accurate. The
information below is subject to change without notice.
ProCurve Feature or Command
Cisco Feature or Command
WAN interfaces named for their technology; E1,

T1, BRI (ISDN), ADSL, etc
WAN interfaces contained under a serial

interface.
Layer 2 interfaces named for their technology;

PPP, Frame Relay, ATM, etc
Layer 2 as part of interface definition.
Interface configuration (use bind command)
Interface and encapsulation contained inside the

interface configuration.
ACL and ACP used together. ACLs may still also

be used alone. ACL and ACP combinations are
ingress only. ACL usage with ip accessgroup command can be ingress and egress.
Traditional ACL usage.
Routed networks use mask to define them.
Routed networks do not require mask.
Appendix D References
These are ranked in order of this papers preference regarding WAN design:
1. Designing Wide Area Networks and Internetworks: A Practical Guide by Marcus, J. Scott:
ISBN: 0201695847
a.
Great text for considering all aspects of a WAN design.
2. T-1 Networking: How to Buy, Install and Use T-1 From Desktop to DS-3 by Flanagan,
William: ISBN: 1578200210
a.
Another classic text and does include some information on E-1.
3. ADSL & DSL Technologies (second edition) by Walter J. Goralski: ISBN: 0072132043
a.
This text also has good all around networking discussions.
4. VPNs (A Beginners Guide) by John Mairs: ISBN: 0072191813

a.
This text also has good all around networking discussions.
5. Practical BGP by White, Russ / McPherson, Danny / Sangli, Srihari: ISBN: 0321127005
a.
Great book for BGP. Easy to read and understand.
6. Wide Area Network Design: Concepts and Tools for Optimization by Robert S.Cahn: ISBN:
1558604588
a.
The depth of this text may not be completely for the intended audience of this
paper so it is prioritized here in this list. This text seems good but was not
critically analyzed to form a more informed opinion.
7. Emerging Communications Technologies, 2nd Edition by Uyless Black: ISBN: 0137428340

a.
A classic reference text.
8. Guide to Designing and Implementing Local and Wide Area Networks, 2nd Edition by Michael
Palmer and Robert Bruce Sinclair: ISBN: 061912122X
a.
Intended to be a course in LAN and WAN Design and Implementation, it spends

10 of its 11 chapters on foundational LAN and WAN material, which is useful,
and 1 chapter on design and although much must be considered from earlier
information given in the text, only 3 pages dedicated to WAN design. I still
recommend the book if one needs a good discussion of both LAN and WAN
technologies and design.
9. Telecommunications Technologies Reference by Bradley Dunsmore and Toby Skandier:

ISBN: 1587050366
a.
A very nice text on telecommunications technologies.
10. The Telecommunications Illustrated Dictionary, 2nd Edition by Julie K. Petersen: ISBN:
084931173X
106
11. Newton's Telecom Dictionary, 21st Updated and Expanded Edition by Harry Newton: ISBN:
157820315
107
J-carrier WAN connections are a closely related variant of T-carrier WAN connections. Like
T1 WAN connections, J1 WAN connections provide a transmission speed of 1.544 Mbps and 24
channels. J1 WAN connections also perform all signaling and control functions in-band.
However, J1 WAN connections use a slightly different framing format for signal transport
than T1 WAN connections use. J1 WAN connections use the J1 signaling format, rather than DS1.
Although the J1 standard is defined, most PTTs in Japan dont appear to be using this
standard for leased lines. For example, Nippon Telegraph and Telephone (NTT), the largest PTT
in Japan, offers a DS3 connection with a transmission speed of 44.736 Mbps.
PTTs in Japan are using the T1 standard because they are delivering WAN connections over fiber optic networks that support
the Synchronous Optical Network (SONET) standard. SONET is the standard adopted by the United States and Canada for
fiber optic networks. Not surprisingly, SONET provides backward compatibility for T-carrier lines.
To find out more about

ProCurve Networking
products and solutions,
visit our web site at
www.procurve.com
2005 Hewlett-Packard Development Company, L.P. The information contained

herein is subject to change without notice. The only warranties for HP products and
services are set forth in the express warranty statements accompanying such
products and services. Nothing herein should be construed as constituting an
additional warranty. HP shall not be liable for technical or editorial errors or
omissions contained herein.
4AA0-0717ENW, 8/2005

Wan Design Guide

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Wan Design Guide

Загружено:

Авторское право:

Доступные форматы

ProCurve Networking by HP

WAN Design Guide

Using ATM and DSL together................................................................................. 36

Advantages and Disadvantages

Summary of Major Points

What to Determine During Planning or

How The Technology is Used

Solution Configuration Examples

Secure WAN Design Overview

Is this a new installation or replacing an existing?

What levels of security need to be in place?

What would be the areas we could compromise?

How much monthly cost do I tolerate for one day of downtime?

Overview of WAN Environments

Figure 1: General Carrier Supported WAN in N.A. Needs figure title

Local Access and Transport

Local Access and Transport

Figure 3: Basic Infrastructure of North America

How is LAN different from WAN?

Types of WAN Circuits

Figure 4: Types of WAN Circuits Illustrated

Dedicated physical circuits

Dedicated Physical Circuits

Switched Physical Circuits

Permanent Virtual Circuits (PVCs)

Switched Virtual Circuits (SVCs)

Figure 5: Simplified WAN Physical Layer Connection Paths

Layer 2 Frame Relay

Figure 6: Frame Relay Connection Paths

How is LAN similar to WAN?

Network protocols and interconnection

architectures in general such as Bridging,

How certain services and infrastructure

work such as their Web Server or email.

What are their requirements for security?

The topological layout of the WAN how

The cost and performance offerings from

various service providers and/or carriers.

Figure 7: Overly Simplified Sketch of Customer Requirements

Figure 8: Basic Representation of WAN Path between Sites

What is the number of clients or servers

on site one that need access to site two?

What is the raw bit rate of the traffic flow

Is this transmission rate different when

Is the full bit rate available for data or do

Is site one primarily comprised of a

single VPN client that requires access to

Designing the Physical and Data Link Layers

The physical transmission media.

generating, transmitting, and receiving

Data-linklayer protocols that provide

logical flow control for moving data

Office Channel Unit

Figure 9: Infrastructure Common to Carrier Local Loops

Summary of Major Points

Both are fixed bandwidth.

How This Technology is Used

It can carry both data and traditional

Multiple T1 or E1 links may be combined

What to Determine During Planning or for Implementation

Channels available for your data, all 24,

Frame Type? For T1 the ProCurve

Secure Router 7000dl series supports D4

Will voice services also be carried on this